Amazon confirmed Wednesday that it laid off an unspecified number of employees in its artificial general intelligence (AGI) organization, the division working on the company’s advanced AI models.

The move, first reported by Reuters, comes as the company invests heavily in programs to help businesses implement AI effectively, including a $1 billion initiative to embed AWS engineers with customers building agentic AI systems.

It’s part of a larger shift in the industry as tech giants and AI frontier labs look to make sure the enormous sums they’re spending on AI pay off in tools businesses actually use.

In a statement, an Amazon spokesperson said building large AI models remains “one of the most important things we’re working on,” but said the company is also “sharpening our focus on the initiatives that matter most for customers, so we can move faster on what counts.”

“That focus means some difficult decisions, including eliminating some roles within parts of our AGI organization, even as we continue to invest in the areas most important to our customers’ future,” the spokesperson said.

It’s the latest in a series of changes in Amazon’s AGI group, which despite its name has always been focused more on frontier models than on what the industry considers AGI, the still-theoretical systems that would match or surpass human intelligence.

Rohit Prasad, the senior executive who oversaw Amazon’s AGI work, left the company late last year, and AGI Lab head David Luan departed in February. In December, Amazon folded the AGI group into a larger organization led by senior vice president Peter DeSantis that also includes chip development and quantum computing.

The cuts are the latest in a series of smaller reductions since January, when Amazon eliminated 16,000 jobs across the company. Amazon said U.S. employees whose jobs are cut will receive 90 days of pay and benefits, outplacement support and transitional health coverage, along with eligibility for severance.

Nearly 20 years ago (!), in 2007, I published my first and only book: Microsoft 2.0. It focused on changes I expected at the company in the “Post-Gates” era. What would remain the same and what likely would be different once co-founder and CEO Bill Gates had left the building?

CEO Satya Nadella has not exited the company (yet). But there’s no question that Microsoft and its mission have morphed considerably in the past year or two. I’m not quite ready to christen this the Microsoft 3.0 era, even though Nadella handed the reins of Microsoft’s dominant commercial business to Judson Althoff nearly a year ago.

That decision resulted in Nadella moving into more of a “founder mode” role, allowing him to focus less on the day-to-day work of running the business. (Microsoft historians may recall that Gates made a somewhat similar move back in 2000 when he became Microsoft’s chief software architect.)

While it might not yet be time for Microsoft 3.0, we arguably could be in the “Microsoft 2.5” era. Windows and Office are still around and still play a big role. Microsoft still builds and sells developer tools and databases. But there’s no question that the cloud and all things AI are at the top of the pecking order now.

I’m embarking on a series here at GeekWire that will focus on what matters to Microsoft and, by extension, to its customers, partners, investors, and employees these days. Who are some of the people shaping and leading the company? What are their opportunities and challenges right now?

Over the next few weeks, I will be profiling various Microsoft execs working on plans for Microsoft’s ongoing evolution. Some are company veterans; some are newcomers. I’ll be talking with top execs from Microsoft’s Security, Copilot, Windows + Devices, Xbox, GitHub, and more.

I’m interested in their strategies for Microsoft’s key products and technologies and how they plan to try to turn Microsoft’s ambitious vision into reality. What are their teams building? What do they see as their biggest challenges and opportunities? And where do they see the technologies in their respective areas heading?

I feel like many of us who’ve been keeping track of the biggest tech companies (myself included) have fallen into the trap of blaming or attributing everything a company does to AI. Layoffs? AI is the culprit. Price increases? It’s all thanks to AI. Changing sales strategies? Chalk it up to AI …

But upon further reflection, I believe Microsoft’s strategy is more nuanced than “AI or bust.” There’s no question that Microsoft’s AI ambitions are shaping its goals and tactics. But Microsoft, as a heavily enterprise-focused entity, can’t simply stop supporting products that aren’t built from the ground up with AI (as much as it might like to do so). Nor can it just leave behind customers who aren’t 100% onboard with its AI moves.

Couple those enterprise hurdles with some not-so-popular consumer decisions, like axing 3,200 people in the gaming unit, and Microsoft’s approach to turning the ship looks a lot trickier.

Our Microsoft 2.5 series kicks off Thursday. Stay tuned.

Jean Paoli has spent his career making documents readable by machines — first as a co-creator of XML, then helping build the file formats behind Microsoft Office. Now his Kirkland, Wash.-based startup, Docugami, is open-sourcing the technology at the heart of its business, betting it can become a standard way to turn documents into data that people and AI agents can trust. 

The company is releasing its technology, called DGML (short for Document Graph Markup Language), under Apache 2.0, a widely used open-source license, so other developers and companies can adopt it.

The idea is to turn it into a shared standard that no single company owns, much as XML became a common foundation across the tech industry. 

The move reflects a shift in where the value is created in AI. Docugami until now has made its money selling software that turns unstructured documents into usable data. It’s betting now that there’s more value in proving that data is trustworthy instead. 

How it works: Docugami is teaming up with Inveniam, a Detroit company whose software helps big investors keep tabs on the mountains of paperwork behind real estate and other hard-to-value assets. Inveniam will record a kind of digital fingerprint of each piece of DGML data on NVNM Chain, its blockchain built with Mantra, a crypto firm that Inveniam is acquiring.

That means, for example, that a single fact buried in a 200-page lease — such as the rental rate, a renewal option, or a default clause — can be verified on its own, without exposing the whole document. An investor, auditor, or AI agent can trace it to the page it came from. 

To work with documents, AI systems usually convert them into a simpler format first. DGML enters a growing field of contenders in that regard, competing with the popular Markdown format and DocLang, a new open standard for AI-ready documents backed by IBM, Nvidia and Red Hat.

The business model: This is a big move for a company of Docugami’s size, taking the 30-person startup in a new direction. Paoli is handing the industry the technology his team spent years building, and pinning the company’s future on a larger idea.

The plan is to make money not from the format itself but from the value of the trusted data. Once a company converts its leases or loans into DGML and anchors the key numbers on the blockchain, investors, lenders and auditors can pay to draw on that verified data.

Docugami will share in the revenue through its partnership with Inveniam. The company also stands to collect a small fee each time a piece of data is recorded on the chain. 

The company is giving away the DGML format and a working version of the software, but not everything. Paoli said the company is keeping some of its own technology private, including AI models it has fine-tuned to read documents, and could sell those or other tools to enterprises. 

“The business model of everybody is changing. And if you know any company where it’s not true, you need to tell me, because I haven’t met them yet,” Paoli said in an interview. 

Docugami has raised about $13 million to date, including a $10 million seed round in 2020 that drew the first investment in Grammarly’s history.

The partnership: Paoli met Patrick O’Meara, Inveniam’s CEO, a few months ago, through a former Microsoft colleague who had become one of O’Meara’s advisers. They quickly realized they had been working toward the same idea from different directions.

Inveniam, founded in 2017, helps big investors keep track of assets that are hard to value, like office towers, private loans and infrastructure. It monitors the documents behind those assets and flags changes as they happen, and its clients include some of the world’s largest sovereign wealth funds, according to O’Meara.

What it lacked was a consistent way to break those documents into verifiable pieces. That is what Docugami provides.

“We’re not putting the data itself on-chain, just a fingerprint of the document. Change one bit, one byte, one pixel, and the hash won’t match,” O’Meara said.

The blockchain comes from Mantra, a crypto company run by John Patrick Mullin. Inveniam invested $20 million in Mantra last year and has since agreed to acquire it outright. Mantra’s OM token collapsed in April 2025, erasing several billion dollars in value. 

Paoli said the project uses the underlying blockchain, not the token.

“Crypto as an industry has gone through a lot of changes in the last 18 to 24 months, and it’s growing up in a lot of ways. This is a real use case with fundamental value, not just pure speculation,” Mantra’s Mullin said in an interview. 

The result is a division of labor: Docugami turns documents into data, Inveniam verifies it and brings the customers, and Mantra provides the chain where the proof is recorded.

The DGML specification, sample documents and reference code are at dgml.io and on GitHub. 

Editor’s note: This story was updated after publication to correct the name of a competing document format, DocLang, and to note that Inveniam’s blockchain is called NVNM Chain.

Two months after emerging from stealth mode, TerraByte AI is using artificial intelligence and a new partnership to upgrade its “Earth Search Engine.”

The startup, which maintains operations in Seattle as well as San Francisco, has just rolled out a TerraByte News service that pinpoints wildfires, earthquakes and severe weather events on an interactive map. Users can follow links to access news reports, social media posts and satellite views related to selected events.

The satellite views include open-source images from NASA’s Earth observation system as well as Europe’s Sentinel satellites. And now the database also features high-resolution pictures provided through a newly announced partnership with Texas-based SkyFi. The partnership gives TerraByte’s users access to SkyFi’s self-service Earth intelligence platform, which offers satellite and aerial imagery from more than 300 sources at prices as low as $15 per image.

“In May, when we came out of stealth, we made the planet searchable,” TerraByte CEO Rishi Madhok told GeekWire. “Now, the moment you find something, you can hold the imagery in your hands within a day. The next step is making Earth intelligence as routine as a web search — you ask, you see, and then you act.”

Madhok and Fuxun Yu, TerraByte’s chief technology officer, founded the company last year as a follow-up to their work on geospatial data analysis at Microsoft. They developed search tools that can recognize features of interest in satellite images and deliver data-driven insights in response to natural-language queries.

Over the past couple of months, TerraByte’s team has grown from three to five employees, Madhok said. “Our goal is to grow the team even further this year, because we are seeing a lot of traction from users since we came out of stealth,” he said.

“A lot of traction is coming from insurance [companies], from the government, from mining, from other areas where there is the possibility to see things,” he said. “And finance, right? A lot of quant firms and hedge funds want to see all of this activity coming in.”

One key application involves emergency response. “Our big focus is on catastrophes, particularly wildfires,” Madhok said. “Our vision is that anybody should be able to track this — not limited to just journalists, but including everyone who is living in those areas and wants to see what’s going on.”

Madhok expects the revenue-sharing partnership with SkyFi to open up new opportunities. “I’m happy to say that we have customers who are paying us,” he said. “From that perspective, we’re already doing well.”

Advances in AI are creating still more opportunities. “Now you can do searches not just using text, but using images, which we call visual search,” he said. “Let’s say you’re searching for a certain kind of vessel, and it’s very hard for you to describe it in natural language. You can just take a screenshot of it, upload it, and within seconds it will literally search for what you were looking for.”

Looking ahead, Madhok and his teammates plan to add people power to the power of AI.

“This is the first version of a platform that we’re going to release, and we obviously want to learn more from our users,” he said. “We want this platform to become crowdsourced, so that people who are local to a region can add more information from that perspective, because then it starts becoming more powerful. We don’t want just TerraByte to be the owner of this.”

Madhok shared a video on LinkedIn that shows how TerraByte’s platform can quickly find high-resolution imagery of a shipwreck in Washington state’s Possession Sound:

Fresh off a $2 billion fundraising and $900 million line of credit, London-based data center startup Nscale is planning a big expansion at a new engineering office in Bellevue, Wash.

Nscale, one of the fastest-growing companies building AI computing infrastructure, recently inked a deal for nearly 24,000 square feet of space at The Eight office tower in downtown Bellevue.

The office is slated to open in January 2027. It will serve as Nscale’s primary engineering hub in the United States, a company spokesperson said. The company currently employs about 50 people in the Seattle area, and the new office will be able to accommodate up to 250 people.

The company earlier this year hired Nidhi Chappell, the former Microsoft corporate vice president who led Azure AI and high-performance computing infrastructure, including the supercomputers that power ChatGPT. As Nscale’s new president of AI infrastructure, based in the Seattle area, Chappell will oversee the company’s global engineering and data center operations.

“I’ve had a front-row seat to some of the biggest moments in AI over the past several years, but one thing has always stood out: the world remembers the breakthroughs, but it’s the people building the infrastructure behind the scenes who make them possible,” Chappell wrote in a LinkedIn post last week announcing the company’s first “onboarding” event in Seattle.

Nscale, which is also preparing to open an office in New York, said it selected Bellevue because of the Seattle region’s concentration of AI infrastructure talent and its proximity to major customers.

Microsoft is one example. Earlier this year, the companies announced an expanded collaboration to deploy Microsoft’s next-generation AI infrastructure across Europe, including large-scale installations of NVIDIA Vera Rubin GPUs in Norway, Portugal and other locations. Nscale said it would be among the first providers outside of Microsoft to deploy the Vera Rubin platform, supporting Microsoft’s growing AI cloud infrastructure.

The new office is the latest sign of Bellevue’s growing role in the AI economy. The Eastside has become a magnet for companies building AI applications and infrastructure, with xAI, OpenAI, Databricks, CoreWeave, Armada, Anduril and others establishing and expanding offices.

AI companies have been giving a boost to the regional office market overall. Claude maker Anthropic, for example, recently announced an expansion of its offices in Dexter Yard in Seattle.

Nscale was founded in 2024. Its $2 billion funding round earlier this year valued the company at $14.6 billion, believed to be the largest Series C financing ever raised by a European technology company. The capital is being used to expand Nscale’s AI cloud platform, GPU infrastructure and data center footprint across North America and Europe.

Its backers include Astra Capital Management, Citadel, Dell, Jane Street, Lenovo, Linden Advisors, Nokia, NVIDIA and Point72.

News of the Nscale office in Bellevue was first reported by the Puget Sound Business Journal.

A coalition of activists and community organizations plans to rally Sunday outside Anduril’s Seattle office, protesting the defense technology company’s development of artificial intelligence-powered military systems and its growing presence in the region.

The demonstration, scheduled for 9:30 a.m. at Anduril’s downtown Seattle office, is being organized by groups including BAYAN Washington, International Coalition for Human Rights in the Philippines and The International League of Peoples’ Struggle. Organizers say the event will highlight concerns about the use of AI in warfare, autonomous weapons systems and the expansion of defense technology companies in Washington state. They expect more than 50 to attend.

“The rally will respond to urgent developments in the expansion of AI weapons companies in Washington State and will expose Anduril as an engine of U.S.-led wars of aggression and a domestic threat to migrant and working class communities,” the organizations said in a statement.

Anduril said it recognizes the right to protest, while defending its work supporting the U.S. military and service members.

“We respect the right to free speech and we understand that protests are a hallmark of democratic expression,” Anduril said in a statement provided to GeekWire. “That said, it is perplexing when people choose to protest a company dedicated to supporting the very military that safeguards those rights.”

The company’s statement continued:

“At Anduril, we’re proud of our role in helping the brave men and women who risk their lives to defend the freedoms that we all enjoy, freedoms that include the right to stand outside and protest our existence. We’ll continue to honor those serving our country, even when others stand in opposition.”

The protest comes as Anduril expands its operations in the Seattle area, including a new maritime manufacturing and testing operation along Seattle’s historic Lake Washington Ship Canal. GeekWire reported earlier this year that the company has taken over the former Foss shipyard, where it is thought to be testing autonomous vessels for the U.S. Navy.

Founded in 2017 by entrepreneur Palmer Luckey, Anduril has become one of the most prominent defense technology companies in the country, developing autonomous aircraft, maritime systems, surveillance technologies and AI-powered software platforms for military and national security customers.

The company’s Seattle expansion has drawn attention because of the region’s long history as a hub for aerospace, maritime engineering, artificial intelligence and advanced manufacturing. The new maritime facility on the south bank of the Ship Canal represents a new chapter for a site with deep roots in Seattle’s shipbuilding history.

In announcing the rally, organizers cited the company’s work on autonomous systems, including underwater and surface vessels, and raised concerns about the role of artificial intelligence in global conflicts.

The groups also pointed to the ongoing Rim of the Pacific (RIMPAC) military exercises, a multinational naval exercise held in and around Hawaii. The exercise runs through July 31 and includes participation from dozens of nations.

Anduril has increasingly positioned itself as a technology company focused on modernizing defense capabilities, arguing that faster adoption of advanced software, autonomy and AI can improve the effectiveness and safety of military operations.

Sunday’s event is expected to include speeches, testimonials and cultural performances from participating community organizations.

The rally adds a new point of public debate around Anduril’s expansion in Seattle, as the company builds out its presence in a region already home to major technology companies, aerospace firms and a growing defense innovation sector.

In addition to the new facility at the Foss shipyard, Anduril operates facilities in downtown Seattle and Bellevue, where it expanded last summer with a lease for 39,851 square feet of space at Skyline Tower.

Anduril also is rapidly expanding its operations in California, where the company is headquartered. And it is building a massive facility just south of Columbus, Ohio, that it dubs Arsenal-1, described by the company as “the future of American defense manufacturing.”

In May, the company raised a $5 billion funding round from Thrive Capital, Andreessen Horowitz and others at a $61 billion valuation.

The past two years have transformed the world of software development, but there’s at least one area that remains largely untouched by artificial intelligence: the operating-system layer inside phones, vehicles, and other connected devices. 

A Seattle startup called logcat.ai has raised $2.55 million to change that.

Co-founded by CEO Varun Chitre and CTO Tarun Vashisth, two engineers with years of experience building device software, logcat.ai is developing a system of AI agents that autonomously hunt down bugs across the kernel, modem, and firmware of devices running Android or Linux.

The pre-seed round was led by Founders’ Co-op, with participation from Act One Ventures, TheFounderVC, Shorewind Capital, Clayoquot Capital, and Alumni Ventures. 

“It’s one of the toughest areas of software engineering, and it doesn’t get a lot of exposure. Operating-system engineering is virtually hidden today,” Chitre said in an interview.

It’s also a challenge for many companies given a shortage of engineers who specialize in the field, compared to the much larger population of developers who build apps and software that run on top of the operating system.

How it works: An engineer using logcat.ai uploads the log files a device generates when something goes wrong — such as bug reports and kernel logs — and logcat.ai’s software analyzes them together to find the root cause and point to where in the code to fix it. Each finding cites the exact log line it came from, so an engineer can check the work.

Currently, logcat.ai finds the root cause and recommends a fix. The larger plan is to have the AI write the fixes, test them, and eventually build new features on its own, with engineers approving the work before it’s deployed.

The long-term goal, Chitre said, is to become the standard tool for building and maintaining operating systems on new and existing hardware — from smartphones to cars to robots and other embedded systems — so a company can ship without a full-stack specialist on staff.

“We’re moving toward a world where software and intelligence extend far beyond our laptops and phones, yet the tooling to build high-quality products for that world is still missing,” said Aviel Ginzburg, general partner at Founders’ Co-op, in a statement.

He called Chitre and Vashisth “one of the only teams in the world truly up for the challenge.”

Traction: The company says it has served hundreds of engineering teams in a public beta, analyzed more than 10 billion lines of trace data, and run thousands of automated investigations. It’s generating revenue but isn’t ready to disclose numbers or customers. 

Competitive landscape: Chitre said logcat.ai’s main competition isn’t another product but in-house scripts and the knowledge locked in a few senior engineers’ heads. App-level crash tools like Google’s Crashlytics and Sentry stop at the app layer and don’t do the deeper system debugging.

Specialist vendors and the contract manufacturers that build devices are potential partners more than rivals, Chitre said, since they face the same engineer shortage.

GeekWire first reported on logcat.ai in March, in a Startup Radar roundup.

The team: Chitre and Vashisth met at Esper, the Bellevue, Wash.-based device-management company, where they worked together for more than seven years. They started logcat.ai because they had spent years doing debugging by hand and knew what was missing.

Chitre has spent more than 13 years in the field, getting operating systems to boot and run on new hardware and porting new Android releases and Linux kernels onto older devices. He was also a maintainer of LineageOS, a widely used open-source version of Android. 

Vashisth has led engineering teams working across Android, Linux, and iOS, and brings a background in large-scale distributed systems. At Esper, he rose to senior software engineering manager. His prior experience includes platform-architecture engineering at Target.

For now, the company is just the two founders: Chitre in the Seattle area, Vashisth in Bengaluru, India. They plan to hire about 10 people over the next year, with a distributed team working remotely from wherever they can find the specialized talent.

They know those hires won’t be easy to find, given the scarcity of people in the field. “That’s the same shortage our product exists to address,” Chitre said, “and we’re not exempt from it.” 

JPMorgan Chase is building out a new AI software infrastructure team, anchored in Seattle, focused on running AI across its data centers and outside providers in a way that controls costs, protects its intellectual property, and avoids tying its fortunes to any one vendor.

Lori Beer, the bank’s global CIO, discussed the effort as part of a broader interview Tuesday during a stop in Seattle. She said the bank is being “careful about lock-in, strategic risk, financial risk, all those things.”

The move comes as business and tech leaders — including Microsoft CEO Satya Nadella and Palantir CEO Alex Karp — publicly warn about the risks of letting a small number of AI vendors accumulate control over costs, data, and the choice of which AI tools businesses can use.

Beer described the new group as an AI infrastructure team but said it works at the software level, separate from JPMorgan groups that build data centers or procure hardware.

She said the group will, for example, develop systems to determine when to route different types of AI workloads to JPMorgan’s own data centers, when to tap into public cloud providers, and when to use newer specialty computing suppliers.

AI agents are one example of where the bank is drawing a line.

Beer said JPMorgan will build and own the software that runs its agents, while treating the underlying AI models as interchangeable. The agentic layer is specific to JPMorgan’s business, whereas the underlying models are general-purpose, and JPMorgan wants to be able to switch among them as the market changes. 

Cost is another focus. Given the option, Beer said, engineers naturally reach for the newest and most powerful model, even when a cheaper one works as well. Systems built by the new team will route specific workloads to different types of models.

The new AI infrastructure team will be spread across multiple JPMorgan locations, but Beer said the Seattle area offers a high concentration of the required skills, including engineers who built cloud infrastructure at Amazon, Microsoft, and other tech platforms before joining JPMorgan. 

It’s part of a broader focus on AI at JPMorgan’s Seattle Tech Center, which has grown to about 400 people since opening in 2018, with a heavy emphasis on cybersecurity.

JPMorgan said this week that it has named Ture Armas, the bank’s CTO for Commercial Bank Lending Technology, to lead the Seattle Tech Center. Armas will continue in his existing role while adding oversight of the tech center’s strategy, talent, and community engagement. He replaces Mamtha Banerjee, who left in March.

The Seattle Tech Center is preparing to move next month into an expanded space at the JPMorganChase Center, the skyscraper that was renamed from the Russell Investments Center in January. The tech center is currently located in a smaller space in a nearby building. The move will put engineers closer to business teams, which Beer called critical as AI accelerates the pace of product development.

Beer, who started her career as a software engineer at a nuclear facility, joined JPMorgan in 2014 from health insurer WellPoint. In 2017, she became the first CIO to sit on the bank’s Operating Committee. She oversees a technology division of about 70,000 people, including 45,000 engineers, with a $20 billion annual budget. 

JPMorgan reported record second-quarter results Tuesday morning, topping Wall Street expectations. On the earnings call, CEO Jamie Dimon said the bank has almost 1,000 AI use cases across the business, with about 50 he described as the most important, in areas including risk, fraud, marketing, note-taking, and document reading.

In what turned out to be a preview of Beer’s comments later in the day, CFO Jeremy Barnum described the bank’s AI priorities: “Use the right model for the right purpose, be smart about open source where appropriate, and ensure that you’re getting value out of it ultimately.” 

As AI workloads drive soaring cloud bills, more companies are weighing whether to move computing out of public clouds and into their own data centers. But building and operating AI infrastructure is far more complicated than simply buying servers — networking has become one of the biggest technical hurdles.

That’s the opportunity Seattle startup Hedgehog is chasing.

Founded in 2022 by CEO Marc Austin, a Cisco networking veteran, Hedgehog develops open-source software designed to make private AI data centers operate more like hyperscale clouds. It has raised $11 million in seed funding, with plans to raise a series A financing round.

We caught up with Austin for the return of GeekWire’s Startup Spotlight to learn more about the 20-person company, the AI networking boom and what surprised him most about building a startup in one of tech’s fastest-moving markets.

In 50 words or less, give us your elevator pitch?

Hedgehog is open-source software that makes AI networking simple. AI clouds and enterprises use it to run GPU networks the way hyperscalers do — deployed in hours instead of months, operated by DevOps teams instead of armies of network engineers, on open hardware with no vendor lock-in.

What problem are you obsessed with solving?

Time to GPU value. A GPU cluster is the most expensive asset most companies will ever buy, and every day it sits idle waiting on the network is money burning. That wait is rarely the hardware — it’s the fabric: weeks or months of scarce network engineers hand-designing, cabling, tuning, and validating it across proprietary CLIs and locked-in vendor gear.

Meanwhile the people told to “own the network” usually aren’t network engineers at all — they’re platform and DevOps teams. We’re obsessed with collapsing that timeline: declare your network like intent in Kubernetes and go from racked GPUs to inference in hours instead of months — on open hardware, no lock-in, no room full of specialists. Cloud-grade networking without hyperscaler headcount.

What surprised you after talking to customers?

How rarely the buyer is a network engineer. It’s platform and DevOps teams, often at AI clouds who just took delivery of thousands of GPUs who are told “you own the network now.” They don’t want to learn BGP; they want a network that behaves like the rest of their cloud-native stack. The other surprise: they don’t just want to run the network, they want to sell it by carving up capacity for their own customers, like a cloud provider does.

How has AI changed the way you build your company?

Twice over.

Our product exists because AI broke traditional networking. Training and inference traffic melts networks designed for web apps.

And AI changed how we build: we use it heavily across engineering, testing, and go-to-market, which lets a small team continuously test every supported device and configuration in our lab and ship with hyperscaler-grade rigor. AI raised the bar for what a startup-sized team can deliver.

What’s one thing people misunderstand about your startup?

That “open source” means hobbyist. The opposite is true: openness is the enterprise feature. Our customers can audit every line of code that runs their fabric, extend it, and never get locked in. Nearly every competitor markets “open networking” while shipping a proprietary controller. Hedgehog is the only one that actually publishes the repo.

What’s the toughest decision you’ve made in the past year?

Betting entirely on Ethernet. We decided open, standards-based Ethernet would win AI networking and put everything behind it. Watching the industry’s largest AI operators now standardize on that same approach makes us feel good about the call — but saying no was hard.

What’s the one piece of advice you give to other entrepreneurs?

Pick the wave, not just the surfboard.

Product decisions are recoverable; betting against a structural industry shift isn’t. Find the standard, the architecture, or the buyer behavior that’s inevitable, align everything to it early, and be patient while the market catches up to your bet.

We’ll know our company has made it when…

Networking is boring again. When a platform engineer stands up a multi-tenant GPU cloud and the network is just a few lines of declared intent that nobody thinks twice about. When “network like a hyperscaler” describes every AI cloud, not just the giants running on Hedgehog, then we will have made it!

Vieu, a Seattle startup aiming to replace cold outreach with warm introductions, launched what it calls the “Business Graph,” a live map of trusted relationships that drive business-to-business sales, marketing, recruiting and fundraising.

The 40-person company, which raised an $11 million seed round in October 2024, has grown to more than 100 enterprise customers including a number of well-known companies. Vieu competes with sales-intelligence tools like ZoomInfo and Outreach, and overlaps with LinkedIn’s Sales Navigator.

The company is led by CEO Samir Manjure and CTO Simon Skaria, both Microsoft alumni. Manjure went on to found KenSci, a healthcare AI startup acquired by Providence in 2021. Skaria has also founded and sold two other startups, Office365Mon and Albits.

The Business Graph, which launched Tuesday, maps relationships between people and companies based on observed signals — such as shared work history, co-authored research, board affiliations, and joint ventures — rather than the self-reported connections that populate LinkedIn.

Common use cases include finding someone who can make an introduction to a decision-maker at a target account, quietly checking references on a job candidate, and figuring out which LinkedIn connections a salesperson actually knows versus the ones they simply accepted a request from.

Vieu says the graph can be used inside its own app or queried directly by AI assistants like Anthropic’s Claude and Google’s Gemini, and it integrates with CRM, email, and Slack.

Manjure said Vieu still has the majority of its 2024 seed round in the bank and has not raised new funding. The company charges customers a platform fee for access to the Business Graph plus outcome-based pricing tied to specific use cases like sales, recruiting, and fundraising.

Sunny Gupta has led two prior enterprise tech companies with backing from venture capital firm Madrona in the past 20 years. iConclude sold to Opsware. Apptio sold to Vista Equity Partners, then to IBM for $4.6 billion.

Now they’re getting the band back together for the AI era. Madrona’s Matt McIlwain is calling it the biggest opportunity “by far.”

Gupta is launching Thira, a Bellevue, Wash.-based enterprise AI startup, with Apptio co-founder Kurt Shintaffer, and leaders from companies such as Atlassian, Oracle, and Databricks. Thira announced Tuesday that it raised $21 million in seed funding led by Madrona, with participation from FUSE.

The idea: Thira is building AI to handle the behind-the-scenes tasks that keep big companies running, like setting up a new hire’s laptop, resetting a locked account, or approving a software purchase. The pitch is to enable a “back-office that runs itself,” according to the company.

It’s starting with IT support. The company is building software agents that can take an IT ticket, work it across the systems where the actual fixes happen — such as ServiceNow, Jira Service Management, Freshservice, and the identity and device-management tools that connect them — and close it out.

Finance and HR systems are also on the roadmap. Thira’s job listings describe agents built to “autonomously run the back-office work that consumes companies today, across IT, finance, HR, and beyond.”

Thira is entering a crowded market. ServiceNow closed its $2.85 billion acquisition of Moveworks last December to build autonomous IT ticket resolution into its service management platform. Startups including Aisera, Rezolve.ai, and Serval are pursuing similar territory.

Part of Thira’s bet is that Gupta and Shintaffer’s relationships with CIOs, which they built over many years at Apptio, will help to give it a foot in the door. Thira says it’s working with 10 companies as design partners ahead of a broader launch this fall.

In many ways, it’s a step beyond Apptio, which helps CIOs see where their companies spend money on technology. Thira is aiming to go past visibility to the “system of execution,” actually doing the work.

In a post on LinkedIn, Gupta said he began hearing from CIOs during Apptio tenure who wanted not only visibility into spending but also the ability to act on inefficiencies and automate work.

“In early 2026, I asked more than twenty CIO friends a simple question: has enough changed that what they’ve been asking for is finally buildable? The answer was yes, and bigger than I expected,” he wrote.

Thira’s team also includes:

• Mudit Goel, previously SVP of engineering at Atlassian;
• Grant Neuman, who was an AI engineer at Oracle Cloud Infrastructure;
• Tarek Madkour, previously director of product management at Databricks;
• and Shashwat Srivastav, previously founder and CEO at AIVOT.

Gupta has been Smartsheet’s executive chair since August 2025, when longtime CEO Mark Mader retired. He also served as acting CEO until Raj Singh was named CEO in October 2025. Shintaffer was Smartsheet’s CFO from July 2025 to May 2026.

McIlwain, the Madrona managing director, is joining Thira’s board of directors. FUSE founding partner Kellan Carter is a board observer.

In a statement, McIlwain said the founding team pairs Gupta and Shintaffer’s two decades of enterprise credibility at Apptio with what he calls “AI-native innovators.” He added, “This is my third time starting and building a company with Sunny and it is by far the largest opportunity we have pursued together.”

Attitudes towards AI differ by country, gender, profession, age, and political affiliation.  A few of those gaps are startling. This article is chock-full of stats. Read it for the surprises, or glance at the bar graph below for a quick overview.

Let’s start with geography, the widest split of all. Ask people in China whether they trust AI and, Edelman finds, nearly nine in 10 say yes; ask Americans and barely a third do. The same chasm shows up, in the Stanford AI Index, on the larger question of whether AI’s benefits outweigh its drawbacks, where most Chinese say it’s good stuff and most Americans have their doubts. 

Here’s a possible explanation. Where economies are young and growing fast, AI reads as a ladder up; where they are mature, it reads as a threat to jobs and more. Trust in AI seems to track two things, confidence in institutions and the expectation of personal gain, and both run higher in many Asian countries than in a wary West.

In the U.S., men are about twice as likely as women to expect AI to be good for society, Pew finds, and the gap is wider still among the researchers who build it. The tempting explanation, that women use the tools less, no longer holds: over the past two years women have drawn even with men in using chatbots, yet they trust them less. Women are also likelier to say AI is moving too fast. 

Adults under 50 reach for ChatGPT at twice the rate of their elders, Pew reports, yet it is the under-30s who are most convinced it will be bad for society. Here, familiarity breeds unease, and for a concrete reason: the young are not only the heaviest users but the most exposed. AI may be coming first for the entry-level jobs they are trying to land, and they sense it, with Gen Z likelier than any older group to expect it to cut into their job prospects, per the Harris Poll. 

Among the AI researchers surveyed, most expect the technology to help the country over the next two decades, Pew’s survey shows; among the public, fewer than one in five do. Some of that is knowledge, since the experts grasp what the systems can and cannot do and fear the lurid scenarios less.

Of course, the people who design AI have their careers and fortunes riding on its success, while the people who answer phones or drive trucks see mainly the threat to their own. The same pattern runs across industries, from technology workers who welcome AI on the job to transportation workers who oppose it. As per Miles’ Law, where you stand depends on where you sit.

The last divide is one that’s moved in recent years, and it’s moved fast. Two years ago Republicans were the AI skeptics; Democrats have since caught up and passed them. Today, just over half of Republicans now trust Washington to regulate AI; barely a third of Democrats do, Pew finds. 

AI companies are now more admired on the right than the left, a Harris Poll shows. Democrats are cooling on companies they once cheered, and Republicans are warming to a boom their side now champions. That said, in both parties more people worry that regulation will do too little than too much; what they split on is whom they trust to do the reining.

Despite some loud voices, there is no single verdict on AI.  Optimism comes from those with the most to gain, in the rising economies and inside the labs; doubts rise from those with the most to lose or the most to fear. Whatever AI turns out to be, it is being built by the people most enthusiastic about it, for a public that is not.

While investors spent much of the spring concerned that frontier AI models from companies like Anthropic and OpenAI would consume the software industry, Dave Clark was closing a funding round for exactly the kind of enterprise software those models are supposedly going to replace.

Auger, the supply chain technology startup founded in Bellevue, Wash., by the former Amazon executive, has raised $50 million in Series B funding led by Eclipse, with existing investor Oak HC/FT also participating in the new round.

The round brings total funding to $150 million for the company, which has grown to about 130 employees and counts Meta’s virtual and augmented reality division, sports merchandise giant Fanatics, and consumer products maker Kimberly-Clark among its customers.

Clark’s view is that general-purpose AI can generate insights but can’t handle deeply specialized domains like running a supply chain. Making financial and operational decisions and executing them at the scale of big companies requires systems built on strong supply chain expertise — what Auger calls its ontology, essentially a detailed map of how supply chains actually work.

“Many a pure technology company died on the hill of supply chain over the last decade,” said Clark, the company’s CEO, in an interview this week. “You really need to understand the complexity and the contextual requirements.”

Auger sits on top of a company’s existing systems — ERP, warehouse management, transportation management, and demand planning tools — and unifies the data into a single operating layer. Rather than replacing those systems, it connects them, using AI agents and traditional optimization models to make decisions and execute them automatically, as much as possible.

For example, in a recent demo at the company’s Bellevue office, Clark showed how the system would handle a supplier missing a delivery commitment when there isn’t enough product to go around. Auger identifies the shortfall, determines which customers get priority, reallocates inventory, and pushes the updated plan back to the company’s existing systems.

Most supply chain software, Clark said, generates alerts and waits for a person to act. Auger is designed to make routine decisions on its own and flag the exceptions for human review.

“We’re not really a tool,” he said. “We’re really the new employee.”

At Fanatics, the sports merchandise company, Clark said about 85% of decisions in the process Auger manages are happening autonomously, with a goal of reaching the mid-90s soon. In addition to the customers it has named so far, Clark said another eight to 10 companies are in contract negotiations or pilot programs.

Clark spent 23 years at Amazon, rising to lead the company’s worldwide operations and later its worldwide consumer business. He left in 2022 and became CEO of Flexport, the freight forwarding startup, but that tenure lasted less than a year amid a turbulent period for the company.

He launched Auger in 2024 with a team that includes Leigh Anne Clark, his wife, who serves as co-founder and president of the company’s fashion and beauty division, focused on an industry Clark describes as one of the most wasteful supply chains outside of groceries.

Clark moved back to the Seattle area from Texas to tap the region’s talent pool, and raised a $100 million Series A from Oak HC/FT. The company quickly assembled a C-suite drawn heavily from Amazon’s senior ranks, along with leaders from Johnson & Johnson, Microsoft, and Salesforce, spanning supply chain operations, AI, data science, and product development.

In March, Auger was named a premier supply chain partner on Microsoft Fabric, the tech giant’s data platform. Auger’s product is built on Azure, and Microsoft sales reps can earn commission on Auger deals. Clark said the partnership has generated engagement but is still early.

Clark said Auger went out for the Series B early, before the company needed it, to avoid the distraction of fundraising during what he expects to be a busy fall of customer onboarding.

With the investment, Eclipse partner Jiten Behl joined the Auger board, which also includes Clark, president and CFO Alex Ceballos, and Oak HC/FT’s Matt Streisfeld.

Auger hasn’t disclosed revenue or other financial metrics, but Clark said the valuation was roughly double the level set by Auger’s initial round. “We didn’t shoot for the crazy astronomical valuation,” he said. “We sat at a place that we felt really comfortable with.”

That pragmatic approach extends to how Auger operates. In Bellevue, the company works out of an office it subleased after Microsoft vacated the space. Auger kept the desks, monitors, and chairs the tech giant left behind, furnishing its new offices for next to nothing.

But Clark’s ambitions for the company are anything but modest. He said Auger’s goal is to have half of U.S. GDP flowing through its platform by 2030, with revenue exceeding $1 billion.

“That requires a pretty steep curve to get there,” he said. “We’re not playing small.”

Rohit Talluri learned the tradition at Amazon: always keep an empty chair in the room to represent the customer — a reminder of the people who will ultimately use whatever gets built.

Now, with AI coding tools creating software faster than ever, Talluri and his co-founders, fellow Amazon veterans Jean Farmer and Gabriel Fong, recognize that the customer can be easily forgotten in the process. So they’re creating a seat at the table for AI agents.

That’s the idea behind Primitive Labs. The startup is building what it calls behavioral intelligence: systems that observe, reason and act as customers would across software platforms and devices, helping product teams learn how people will react to a new feature, design or marketing decision before it ships.

Traditional user research and focus groups can take weeks or months, so teams under pressure to ship quickly are tempted to skip them. Primitive Labs is automating that research with agents that simulate human behavior, aiming to make it a routine step in building software.

“It’s bringing humans back to the center of a world that’s created by AI,” Talluri said. “That is the goal here.”

The mission, according to the startup’s launch post, is to “make human behavior a first-class primitive of software development.” That’s the inspiration for Primitive Labs’ name. The idea is to build products that people will understand, trust and keep using — not the average user, but specific types of users in specific contexts.

Founding team: Talluri, the Primitive Labs CEO, is joined by co-founders Farmer, CTO; and Fong, COO.

Fong and Talluri have worked together since 2020. At AWS in Seattle, Fong held product marketing and enterprise account roles, then led sales and marketing at the cloud consultancy DoiT International.

At Primitive Labs, his role runs broader than sales and marketing, spanning product direction, customer development and operations. Talluri describes him as highly technical and a hands-on contributor to the company’s core product work.

Farmer and Talluri worked together at AWS on large-scale machine-learning infrastructure, including the SageMaker HyperPod training service, before both moved into Amazon’s AGI organization.

Farmer worked on the Amazon Nova models’ ability to use software tools — designing how the models call tools and take actions, and building the systems to test and measure how well the resulting agents perform. That work included benchmarks for the Model Context Protocol (MCP), the emerging standard for connecting AI models to outside tools and data.

Roots in AI autonomy: Talluri joined the AGI Autonomy Lab, the group Amazon assembled around talent it hired from Adept, a San Francisco startup building AI agents that operate software on their own.

Amazon had brought on Adept’s CEO, David Luan, a former OpenAI executive, along with other co-founders in 2024, and licensed the startup’s technology, putting Luan in charge of the lab. Talluri worked there on computer-use agents and helped launch Nova Act, Amazon’s agentic computer-use model.

Talluri said he initially came close to leaving Amazon in 2025 to start a company, before leaders there steered him toward the Autonomy Lab to work under Luan (who has since left Amazon).

Funding: Primitive Labs has raised a pre-seed round, led by a16z Speedrun and joined by several small, newer venture funds and a group of angel investors. The company isn’t disclosing the funding amount.

Its launch post lists backers including Olive Tree Capital, Cloverfield Fund and Unexpected Investments (from former TechCrunch editor Josh Constine), plus angels such as Luan, Harsh Patel and Artur Kiulian, and others with backgrounds at OpenAI, Amazon, Google DeepMind, Databricks, Nvidia and Meta.

Primitive Labs will join a16z Speedrun’s cohort starting this month, and expects to raise its next round around the end of the program, in September or October.

Headquarters: The company is based in San Francisco, where it’s working part-time out of a16z’s Speedrun space, with plans to get its own office after making its first hires.

Talluri, a University of Washington graduate who read GeekWire as a student and dreamed of launching a startup of his own, said the choice came down to San Francisco’s talent density and the pace of AI research there, plus the Speedrun program being there.

Primitive Labs posted its first job listings last week — for founding engineers, researchers and an intern, in San Francisco or New York.

Product status: The company is pre-revenue and working with a small group of early customers who are testing its product and helping shape it, including private previews with what Talluri described as Fortune 500 and Fortune 50 consumer-technology and e-commerce brands.

The company plans to launch its products in general availability later this year.

How it works: The agents work across devices including computers and phones, focused for now on digital products and customer journeys. The company says it has also explored using them to gauge reactions to physical products, such as brand and packaging.

The underlying research draws on computational cognitive science, continual learning and custom memory systems modeled on how people store information — work Talluri said the company plans to publish and partly open-source in the coming months.

While other startups are working on agent-based simulation and automated testing of user interfaces, what sets Primitive Labs apart, Talluri said, is the focus on human alignment. That means building agents that faithfully represent a specific product’s users, and making that a standard layer of how software gets built. He described the key measure as behavioral fidelity, or how closely an agent’s choices track human decisions.

Asked whether the startup will keep a chair empty when it gets an office, in the Amazon tradition, Talluri didn’t hesitate. “100%,” he said. And yes, he said, they’ll be envisioning an agent sitting there.

Microsoft is cutting 4,800 jobs, just over 2% of its global workforce, citing a need to revamp its sales and consulting division to keep pace with a rapidly changing tech industry, while overhauling its Xbox business in a push for long-term growth and profitability from gaming. 

The cuts include about 600 jobs in Washington state, home to Microsoft’s Redmond headquarters. That’s down from 3,200 job reductions locally a year ago. Combined with ongoing hiring, Microsoft’s workforce in the state is expected to remain stable at around 52,000 people.

About 1,600 of the 4,800 job cuts being announced Monday are in the Xbox division. Additional Xbox layoffs in the months ahead are expected to bring total job reductions in the gaming division to roughly 3,200, or about 20% of the global Xbox workforce, this fiscal year. 

Microsoft is also spinning off four Xbox game studios to operate independently. 

In an internal memo, Xbox CEO Asha Sharma called it the biggest restructuring in Xbox history, saying the division has been “operating at margins that are 3-10x lower than comparable platform and publishing businesses” and that studios have been losing 64 cents for every dollar invested.

Overall, top executives sought to distinguish Microsoft from other tech giants, saying the cuts were minimized by the redeployment of more than 4,000 employees into new roles over the past year and a voluntary retirement program that let thousands more exit by their own choice.

By comparison, the company last year cut more than 15,000 jobs globally in two rounds of layoffs in spring and summer 2025 — the largest reductions in more than a decade.

The latest cuts come amid record capital spending on the company’s AI infrastructure, pressure from Wall Street to keep operating expenses in check, and a 30% stock slide that has wiped out roughly $1.2 trillion in Microsoft’s market value over the past nine months.

“Microsoft can only be a strong employer if it has a successful business,” said Brad Smith, its president and vice chair, in an interview with GeekWire. “We have to adapt to change.”

Before the latest cuts, the company’s total workforce was about 220,000 people. Across the company, Microsoft expects worldwide headcount to decline year-over-year, CFO Amy Hood said on an April earnings call. 

Amy Coleman, Microsoft’s chief people officer, said in a memo to employees Monday morning that the roles the company is eliminating today are not being directly replaced by AI.

At the same time, she acknowledged, “AI is changing how work gets done.” She added, “Some of the tasks we do every day can now be automated, and that means we all need to keep learning, keep building new skills, and keep adapting as the work evolves.”

However, the line from Coleman’s memo that may get the most attention internally is this: “We are still early on this journey, and there will be more changes ahead; other parts of our business will need to make similar changes.”

In an interview, Coleman stopped short of signaling further layoffs across the company. Instead, she described a larger shift in how Microsoft manages its workforce. That includes reskilling engineers for customer-facing and AI-focused positions, and exploring how to make voluntary exit programs a regular part of the company’s operations — not just a one-time offer, but potentially something employees could opt into annually or on an ongoing basis.

Coleman confirmed that about 30% of roughly 8,750 eligible U.S. employees accepted Microsoft’s first-ever voluntary retirement program in recent weeks, in line with the company’s expectations, which reduced the size of the reduction in force announced Monday. 

The cutbacks and changes in the company’s sales and consulting teams build on last week’s launch of the Microsoft Frontier Company, a $2.5 billion initiative to embed 6,000 engineers inside customers to deploy AI. The shift is reducing some traditional sales and consulting roles and resulting in more technical positions working directly with customers. 

“We’re seeing that we need more engineering excellence in the customer space,” she said. 

Smith said software development is undergoing its biggest shift in the more than 50 years since Microsoft’s founding. The widespread use of AI is making code cheaper and faster to produce, but he said that’s also creating demand for new kinds of roles and work.

“Some things like coding require less time of software developers,” he said. “At the same time, there’s new parts that are growing, whether it’s the product management or software design, or perhaps most importantly, working directly with customers.”

Update: A filing by Microsoft on Monday under the Washington state Worker Adjustment and Retraining Notification Act listed 605 positions being eliminated in Washington state.

The roles span software engineering, product management, sales strategy, data science, business program management, marketing, and game design, among others — ranging from mid-level individual contributors to senior managers, consistent with cuts that reach across both the company’s technical ranks and its sales and consulting operations.

As the country marks its 250th birthday this week, Microsoft is rolling out an unlikely summer project: a six-part series of short videos, hosted by Microsoft President and Vice Chair Brad Smith, that look to American history for lessons relevant to technology and innovation today.

The premise is that every technology debate of the moment — over such issues as patents, privacy, and who gets to shape AI — has a precedent somewhere in the country’s past, and that we’d all benefit from remembering how we got here in the first place.

“We felt that the 250th anniversary of the country deserved some added reflection about the lessons of history, the role of technology, and the questions that we’re facing as a country,” explained Smith, a well-known history buff, in an interview with GeekWire this week.

In the first episode, for example, he stands in Philadelphia’s Independence Square to explain how a steamboat demonstration on the Delaware River in 1787 helped inspire the Constitutional Convention to give Congress the power to grant patents. This was the basis for the intellectual property framework that Smith describes as a bedrock of American innovation.

Savvy viewers may see some irony in a company extolling the virtues of IP protections even as Microsoft and OpenAI defend themselves against a New York Times copyright suit over the material used to train their AI models.

Asked about that, Smith made it clear he doesn’t see a contradiction.

“Every generation of technology has required a new round of legal thinking, legislation and oftentimes lawsuits, so that courts can sustain the balance that has always been needed between new innovation and the protection of things created already,” he said.

He also noted that Microsoft is often the party going to court to protect customers, pointing as one example to the company’s move this week to intervene before Europe’s top court in defense of the European Union and U.S. data-protection framework.

The six-part series was overseen by Smith’s longtime chief of staff, Carol Ann Browne, a Microsoft vice president; and produced by Kirkland, Wash.-based Trifilm. The episodes, around 3 or 4 minutes each, will roll out in the coming weeks. Smith said they recorded during existing travel plans, working the shoots into stops on trips he was already taking.

The series travels next to a Boston courtroom for the birth of privacy rights, Henry Ford’s Detroit assembly line for the spread of new technology, Cincinnati for Tocqueville’s take on nonprofits, Great Falls, Md., for George Washington’s early infrastructure ambitions, and the Lewis and Clark expedition in Montana for the value of uniting competing viewpoints.

“The 250th anniversary of the country is quite rightly an occasion to honor the past, celebrate the past,” Smith said, explaining the motivation for the series. “But let’s make sure we get something out of the past that helps us be more successful in the future.”

KredosAI, a Seattle-area startup that uses AI and behavioral science to help companies chase down late consumer payments, raised $7 million in a new funding round led by BMW i Ventures, the independent venture capital arm of automaker BMW Group.

The company, founded in 2021 by former T-Mobile executives Balaji Sridharan and Dave Thoms, is based in Issaquah, Wash. It focuses on the period after a bill is overdue but before the account gets sent to collections or written off. Its technology is able to tailor the wording, timing and channel of each overdue message based on a customer’s account history.

The premise, Sridharan said, is that most people aren’t being nefarious in their tardiness but are dealing with something more mundane, such as a forgotten due date, a short-term cash crunch, or possibly some kind of frustration with the service. 

“The majority of consumers who go late on payment actually want to pay,” he said. “There’s a very small subset of people that are fraudsters, but most of them want to pay.”

New investors Motley Fool Ventures and Walter Ventures joined existing backers Okapi Venture Capital, StartFast Ventures, SaaS Ventures and Stout Street Capital in the Series A round. Total funding to date for the company is a little over $10 million.

The BMW connection came through an introduction from an existing investor, Sridharan said. Having an automaker’s venture arm behind it matters, he added, as KredosAI moves deeper into auto lending.

Subprime auto-loan delinquencies have climbed to their highest levels since the 1990s. Lenders, Sridharan said, weigh the problem much the way telecoms do — balancing the cost of recovering a payment against the value of keeping the customer. That overlap, along with BMW’s footprint in the car business, made its venture arm a logical fit. 

KredosAI works with large enterprises, including some in the Fortune 50, though it doesn’t name most of them publicly. It got its start in telecom, which speaks to its roots: Sridharan and Thoms met at Bellevue-based T-Mobile. Sridharan spent eight years there, first running corporate strategy and later the carrier’s IoT unit, following an earlier stint at McKinsey. Thoms has spent much of his career in credit and collections at telecom and financial-services firms. 

Watching T-Mobile wrestle with millions of past-due accounts each month, they came to think there was a better way to handle the conversation with a customer who’d fallen behind. 

To decide what to send, the software weighs a customer’s account characteristics (how often they’ve been late before, their average balance, how long they’ve been a customer) while steering clear of off-limits signals like age. It reaches people through text, email and most recently RCS, along with AI voice agents the company began adding over the past year. 

The company says the approach delivers notable improvement: across its customers, it reports cutting write-offs by 11.5% and lifting customer lifetime value by 13.6% compared with conventional collections. It says its platform has handled more than 200 million customer interactions over the past two years, with revenue growing more than sixfold in that span. 

KredosAI is also a partner of FICO — the analytics firm best known for the FICO credit score — and integrates its technology into the FICO Platform, the software banks and other large companies use for credit decisions and collections.

The company competes with a range of collections-software players, including larger, more established Symend, a Calgary-based company that also uses behavioral science to interact with late-paying customers. The field also includes online debt collectors and companies selling older collections software. 

The company has about 25 employees, roughly eight of them in the Seattle area. 

Sridharan said the funding will go toward sales and marketing, further product development around agentic AI and voice agents, and eventually international expansion. He expects to roughly double headcount over the next year, to 50 people or more. 

The additional funding, he said, “gives us a bit of fuel to go to market a little more aggressively than we have in the past.”

Microsoft is launching a new AI “company.” It won’t be a separate legal entity, and most of its 6,000 people already work at Microsoft. But the $2.5 billion behind it is real, and the stakes are big, given how many of its AI partners and rivals are racing to do basically the same thing. 

The tech giant on Thursday announced “The Microsoft Frontier Company,” which will embed engineers inside customers to build and run AI systems. It will be led by Rodrigo Kede Lima, a longtime Microsoft sales and enterprise leader, most recently president of Microsoft Asia.

This practice is known in the industry as forward-deployed engineering, in which a company sends its own technical employees to work inside a customer’s operations to design, build, deploy and operate AI systems on-site rather than selling a tool and walking away. 

The model was pioneered two decades ago by Palantir, but in recent months the approach has become the hot new thing in enterprise AI. Amazon committed $1 billion to its own forward-deployed engineering initiative just two days ago. (Some inside Microsoft suspect that its rival may have caught wind of what it was planning and moved to announce first.) 

Anthropic and OpenAI launched rival ventures in May to put engineers inside enterprise customers. Unlike Microsoft’s initiative, the OpenAI Deployment Company, as the ChatGPT maker’s venture is known, is an actual standalone entity — majority-owned by OpenAI but backed by more than $4 billion from a partnership led by the private-equity firm TPG. 

Similarly, Anthropic teamed with Goldman Sachs, Blackstone and Hellman & Friedman on a $1.5 billion venture — not yet named — to embed engineers inside mid-sized companies, starting with the investment firms’ own portfolio businesses.

Microsoft is attempting to one-up them all. 

“This goes beyond what has been labeled as Forward Deployed Engineering (FDE) and will be the largest, most capable, outcome-driven engineering organization in the industry,” wrote Judson Althoff, CEO of Microsoft’s commercial business, in a post announcing the new initiative Thursday morning.

Responding to questions from GeekWire, a Microsoft spokesperson called the new initiative “a purpose-built company with its own leadership and financial accountability” but stopped short of calling it a separate legal entity or standalone company.

The spokesperson said the organization “brings together more than 6,000 industry, engineering and AI professionals, drawn primarily from Microsoft’s existing engineering and forward-deployed teams,” noting that it will “grow through a combination of internal talent and external hiring across engineering, AI, and industry roles.”

Separately, some consulting roles are among those expected to be impacted by the round of layoffs anticipated next week.

Microsoft wouldn’t say whether the $2.5 billion is new spending or repurposed from existing budgets, or over what period it’s being spent. The company also hasn’t yet spelled out what the new organization means for the future of its existing consulting and services units.

Across the industry, this is happening now because the payoff from AI has proven harder to capture than many companies expected. Businesses across the economy have adopted tools like ChatGPT, Claude, Gemini and Copilot, only to find that impressive demos don’t automatically translate into results. The technology is powerful, but deploying it can be difficult inside a real company, with its own data, rules and entrenched ways of working.

So the AI providers have started sending their own engineers to work inside those companies, figuring out where the AI can actually help, then building it into their operations.

“Having the model alone doesn’t change your workflows or how you operate,” said Marc Nachmann, Goldman Sachs’ global head of asset and wealth management, in an interview with CNBC about the Anthropic partnership. “You need people who can combine the technology with what’s actually happening in the business and implement those changes.” 

The big AI providers have multiple reasons to do this. Each of them wants to get more businesses using its AI platform at higher volumes. All of them are looking to drive long-term demand for the AI capacity they’re collectively spending hundreds of billions of dollars to build.

Another big reason: AI models are becoming commodities, getting cheaper and more similar by the month. The big money for the likes of Microsoft is in selling the services needed to make AI pay off inside a company, which is a far bigger market than just selling the models themselves.

Microsoft is pitching privacy and trust as a selling point. Its promise is that a customer’s data and hard-won knowledge stay the customer’s alone. Microsoft says it won’t feed them into training its AI models in ways that would hand the same advantages to the customer’s rivals. 

It’s also promising choice: customers can run whichever AI model fits the job, from OpenAI, Anthropic, Microsoft, or open-source providers, not locked into using one.

Microsoft CEO Satya Nadella has argued that a company should be able to exchange one AI model for another without losing all the institutional knowledge it has built up. 

That’s his test, as he put it, for whether a business still controls its own future.

“The last thing any of us want is a world where every company across every sector is ceding value to a few models that eat everything they see,” Nadella wrote in a June 14 essay. “If all the value is accrued by only a few models, the political economy will simply not tolerate it. There is no societal permission for an AI future that hollows out entire industries.”

Whether that vision of swappable AI models becomes a reality remains to be seen. There’s actually a risk for customers that the opposite will happen in the forward deployed engineering approach. Even if they can theoretically swap in a competitor’s AI model, working with Microsoft’s engineers means their systems naturally end up running on Microsoft’s cloud platform and related technologies, making it very difficult to jump ship.

It’s also not clear how new all of this really is for the company. Microsoft already runs a large in-house delivery arm — Industry Solutions Delivery, the group that absorbed what used to be called Microsoft Consulting Services — with thousands of consultants and engineers building and deploying technology inside customer organizations. 

Microsoft also has programs like FastTrack to help customers roll out its software, and over the past year it has been rolling out “forward-deployed engineering” teams with partners, including a dedicated practice with Accenture and a $1 billion, five-year alliance with EY.

So ultimately the Microsoft Frontier Company is less a new company than a new push behind work the actual company was already doing, albeit bigger and better-branded than before.

Venice.ai, a privacy-focused AI startup with strong Seattle ties, has raised $65 million in its first outside funding, valuing the 2-year-old company at $1 billion. 

The company positions itself as a private and unrestricted alternative to mainstream AI services, offering access to a range of open-source and commercial AI models. Venice says it doesn’t log or store users’ prompts and responses on its servers, keeping conversations on people’s own devices. It also strips out many of the content filters built into competing tools. 

The Series A round, announced Wednesday morning, was led by Dragonfly, a crypto-focused investment firm, with participation from North Island Ventures, Coinbase Ventures, Archetype, Morgan Creek, Liquid2 Ventures and Seattle-based Founders’ Co-op. 

The company was founded in 2024 by crypto entrepreneur Erik Voorhees, its CEO. Voorhees founded the crypto exchange ShapeShift and has long argued against heavy government regulation of cryptocurrency.

Seattle tech veteran and serial entrepreneur Jesse Proudman is Venice’s president, CTO and co-founder. The two met as classmates at the University of Puget Sound in Tacoma.

“We want Venice to be thought of in the consumer landscape on the same terms as a ChatGPT or an Anthropic,” Proudman said in an interview. “We want people to open their phones and have our app sitting alongside those apps.”

The case for privacy comes from how people are starting to use AI. As chatbots become go-to tools for sensitive matters — medical questions, legal issues, job negotiations, relationship advice — users hand over intimate details that accumulate in the databases of companies like OpenAI and Anthropic. 

That data, Proudman said, is only as safe as the company holding it.

“It only takes one breach, one disgruntled employee who is going through that data, a government subpoena, a change in government policy — and then all of that data no longer is private to you,” he said. “It can be health records, it can be legal questions, it can be job negotiations, it can be relationship advice.”

Venice’s answer is to create no central trove to breach or subpoena in the first place.

Marketing AI with fewer restrictions can make Venice more useful in some cases, but it also raises the misuse questions that lead mainstream services to build in guardrails in the first place. Proudman said Venice includes some safeguards to prevent abuse and illegal activity. 

The company nonetheless bills itself as an “AI safety company,” casting the surveillance of users’ thoughts — rather than the content of their prompts — as the greater danger. 

Proudman is based in Seattle, where he has spent more than two decades starting and selling technology companies. He founded cloud-computing company Blue Box, which IBM acquired in 2015, and crypto trading startup Strix Leviathan, acquired by hedge fund Parataxis in early 2025. Strix spun out Makara, a crypto investing startup, in 2021, and Betterment acquired Makara the following year. 

Proudman spent about three years as a VP at Betterment, where he started moonlighting on Venice in 2024 — building it nights and weekends before leaving to go full-time.

Venice says it reached 3 million users in April and turned profitable in the first quarter. 

“That hockey stick that we always hear about, and that I’ve spent 25 years trying to build companies to find, finally manifested,” Proudman said. 

Venice makes money through consumer subscriptions and paid access to its developer API. It also has its own cryptocurrency, the VVV token, which developers can buy and lock up to reserve a share of the company’s computing capacity instead of paying per use.

Proudman said Venice will use the funding to build its own data center infrastructure — owning the GPUs that power its service rather than renting computing capacity — and to invest in growth as it tries to establish itself as a mainstream consumer brand. 

The company has grown to about 45 employees, up from roughly 15 people a year ago, with six in Seattle. It operates as a remote team and doesn’t currently have an office. 

Whether Venice expands its Seattle footprint long-term may hinge on state politics. Proudman has publicly opposed Washington’s new 9.9% “millionaires tax” — a state income tax on household income above $1 million that was signed into law in March and takes effect in 2028 — and said he won’t stay in the state if it does. 

He’s pinning his hopes on a repeal campaign that backers are trying to get on the November ballot. 

“I love it here … Seattle is a unique and phenomenal place to build a company, and I’ve been building companies here my entire life,” Proudman said. “I want to see us continue to be competitive against the Bay Area.” 

Anthropic is embarking on a major expansion in Seattle, underscoring how artificial intelligence companies are emerging as one of the few bright spots in the region’s office market.

The maker of the Claude AI model recently finalized a lease at Dexter Yard North in Seattle’s South Lake Union neighborhood, capping months of speculation about the company’s expansion plans in the region.

Terms of the deal were not publicly disclosed, but CoStar News reports that the company leased 113,000 square feet of space across multiple floors in the north tower at 700 Dexter Avenue North. CoStar called it one of the largest office deals of the year so far in Seattle.

The expansion would significantly increase Anthropic’s footprint in Seattle, where the San Francisco-based company established an engineering office in 2024 as it recruited talent from the region’s deep pool of AI researchers and software engineers.

It would also place Anthropic next door to Amazon. The companies in April expanded their existing partnership: Amazon committed to invest up to $25 billion in Anthropic, which simultaneously made a $100 billion-plus spending commitment to AWS over 10 years. 

The following month, Anthropic announced $65 billion in funding at a $965 billion valuation, thought to be the last venture round before an initial public offering later this year.

On Thursday, Anthropic released Claude Sonnet 5, which the company says “can make plans, use tools like browsers and terminals, and run autonomously at a level that, just a few months ago, required larger and more expensive models.”

Also this week, The U.S. Department of Commerce removed export controls on the company’s Claude Fable 5 and Mythos 5 models, part of an ongoing back-and-forth with the Trump administration.

Anthropic’s Seattle lease provides hope that demand from AI companies could help revive parts of Seattle’s office market after several years of elevated vacancy driven by remote work and tech industry cutbacks. Seattle’s office vacancy rate inched up to 28% during the first quarter, the highest in the region.

Other AI firms, including OpenAI and Databricks, have also expanded their Seattle-area office footprints in recent months. In those instances, the companies chose to grow in nearby Bellevue.

Dexter Yard, a two-building office and life sciences campus developed by BioMed Realty, opened in 2022 and was designed to accommodate both technology and biotech tenants. The north tower contains approximately 163,000 square feet of office and lab space.

Anthropic has a number of open engineering roles spread across Seattle, New York and San Francisco. The company says it expects all staff to be in one of their offices at least 25% of the time.

A spokesperson for Anthropic acknowledged the new lease, but did not respond to requests for additional comment.