Weekly Download #34

Hi !

This week in Silicon Valley, startups & tech:

• Capital One Acquires Brex for $5.15B: The fintech darling gets swallowed by a banking giant in cash and stock

• World Labs Hits $5B Valuation: Fei-Fei Li's spatial AI startup 5x'd in less than a year

• Big Tech's $109M Lobbying Record: Meta, Amazon, and Google spent more on DC influence than ever before

• The AI Jobs Squeeze: Amazon targeting 30K cuts, Autodesk slashing 7%, and entry-level hiring is freezing over

• AI Funding Frenzy: Inferact, LiveKit, and Neurophos all raised $100M+ rounds

As we continue to build at Context Ventures, we've launched Startup Intros with a simple mission: to help early-stage founders find the right investors, faster and smarter.

🔥 Context & Startup Intros Events Coming Up

• Jan 28: Agentic AI Night-Out: Join us for a fireside chat with founders who hit 7-figure ARR in their first year, plus networking. Register on Luma!

• May 27: Early Stage Summit: We've rescheduled to make it better than ever. Register on Luma!

💳 Capital One Swallows Brex: The $5.15B Fintech Consolidation

The fintech darling that once symbolized startup disruption is now part of a banking giant.

💰 The Deal Structure

Reuters reports that Capital One has agreed to acquire Brex for $5.15B in a 50/50 cash and stock deal. Brex specializes in tech that administers corporate credit cards, expenses, and rewards, which is exactly the infrastructure Capital One needs to compete in the business banking wars.

The deal is expected to close in mid-2026, pending regulatory approval. For Capital One, this is the latest in a string of acquisitions following its pending $35B Discover deal announced in 2024.

The Valuation Haircut: Brex was valued at $12.3B at its 2022 peak. This deal represents a 58% haircut from that high-water mark, but also a soft landing compared to many pandemic-era fintech valuations that went to zero. Brex's investors are receiving liquidity at a time when many fintech exits have dried up.

🏦 Why Capital One Wants Brex

Capital One has been aggressively expanding its commercial banking presence, and building expense management software from scratch would take years. Brex brings a turnkey solution:

The Product Stack:

• Corporate card infrastructure used by tens of thousands of startups and growth companies

• AI-powered expense management that automatically categorizes and reconciles transactions

• Rewards programs tailored to business spending patterns

• Bill pay and accounts payable automation

The Customer Base: Brex's sweet spot has always been high-growth startups or companies that start small but scale quickly. For Capital One, these relationships are a pipeline to future commercial banking clients. A company using Brex for expense management today could be a $100M credit facility customer in five years.

Translation: When a bank pays $5B+ for a fintech, it's not buying the product; it's buying the distribution. Brex's relationships with startups and growth companies are worth more than its software. The acqui-hire era for fintech is officially here, and the message to other fintechs is clear: if you can't IPO, get acquired while you still can.

🚀 CodeLaunch USA - Applications Close in 5 Days

Most accelerators want your equity. CodeLaunch wants your success.

Apply for free by February 1st, keep 100% of your company, and get matched with a professional software development team to build your MVP or advance your product roadmap. Finalists pitch live in Dallas on April 15 to an audience of angel investors, VCs, and entrepreneurs. The winner advances to the World Championship in November.

126 startups accelerated. $64M+ raised by alumni. Zero equity surrendered.

Apply free at codelaunch.com!

🌐 World Labs Hits $5B: Fei-Fei Li's Spatial AI Bet

The AI legend's startup just 5x'd its valuation in under a year, and the tech explains why.

📈 The Raise

Bloomberg reports that World Labs, founded by Stanford AI pioneer Fei-Fei Li, is in discussions with investors to raise hundreds of millions at a $5B valuation. For context, World Labs was valued at just $1B when it emerged from stealth in 2024.

That's a 5x jump in under 18 months, which is one of the fastest valuation climbs in AI startup history. The round is expected to include both existing investors and new strategic partners interested in World Labs' unique approach to AI.

🧠 What Spatial Intelligence Actually Means

World Labs focuses on "spatial intelligence," or AI that can understand and interact with 3D environments. Current AI models excel at processing text and 2D images, but they fundamentally don't understand how the physical world works.

The Applications:

• Robotics: Robots need to understand physical space to navigate and manipulate objects

• AR/VR: Immersive experiences require AI that understands 3D environments

• Autonomous vehicles: Self-driving cars need spatial reasoning beyond simple object detection

• Gaming and simulation: Creating realistic virtual worlds requires understanding how physical spaces work

🎯 The Fei-Fei Factor

Li created ImageNet, the dataset that sparked the deep learning revolution in 2012. Her work is directly responsible for the AI boom we're living through now. When she starts a company, investors pay attention, and a 5x valuation jump suggests they like what they're seeing.

The Team: World Labs has assembled a roster of Stanford AI researchers and industry veterans. When you combine Fei-Fei Li's reputation, a differentiated technical approach, and a team with deep expertise, the $5B valuation starts to make sense.

Translation: The AI infrastructure layer is getting crowded with companies building better language models. Smart founders are moving up the stack to solve different problems. Spatial intelligence is a differentiated bet on what AI needs to do next, not just what it can do now.

🏛️ Big Tech's $109M Lobbying Blitz: Buying Influence at Record Scale

The companies building AI are also building influence, at unprecedented scale.

💵 The Numbers That Matter

Bloomberg analysis reveals the top US tech companies spent $109M on DC lobbying in 2025, topping $100M for the first time in history.

The Leaderboard:

• Meta: $26.29M (largest spender)

• Amazon: $17.78M

• Google: $13.1M

For context, $109M is more than most startups raise in their entire lifetime. When companies spend more on lobbyists than some Series B rounds, the message is clear: policy is now a competitive moat.

🎯 What $109M Buys You

This spending surge comes as Big Tech faces antitrust scrutiny, AI regulation debates, and a new administration signaling openness to deals. The lobbying isn't defensive; it's offensive.

The Shopping List:

• AI regulation (or lack thereof): The biggest AI labs want to shape rules before they're written, not react after

• Antitrust relief: Meta's FTC case, Google's search monopoly ruling, and Amazon's marketplace practices are all in play

• Data privacy frameworks: Rules that don't disrupt existing business models

• Immigration policy: H-1B visa expansion for talent acquisition

• Chip export controls: Influencing which semiconductors can be sold to China

The ROI Math: Consider Google's search monopoly case. The company faces potential remedies worth tens of billions in revenue. Spending $13M to influence that outcome isn't an expense; it's an investment with potentially 1000x returns.

🔮 The Regulatory Runway

Translation: The companies building AI are also building influence. $109M buys a lot of access, and potentially a lot of runway before meaningful regulation arrives. When your lobbying budget exceeds most startups' total funding, you're not just playing the game; you're writing the rules.

The timing isn't coincidental. With AI capabilities advancing faster than regulators can understand them, every month of delayed oversight is another month of unconstrained growth. Big Tech isn't betting against regulation; they're betting they can shape it.

🪓 The Great AI Jobs Squeeze: Layoffs Up, Entry-Level Down

The AI productivity gains executives keep promising? They're starting to show up in headcount, and the pain isn't evenly distributed.

📉 The Layoff Wave Continues

The job cuts keep coming, and the justifications are getting more explicit about AI's role.

Amazon's 30K Target: Reuters reports Amazon is planning a second round of job cuts next week as part of its goal to trim 30,000 corporate workers total. The company already cut 14,000 jobs in October. That's 30,000 corporate roles, not warehouse workers, being eliminated as Amazon restructures around AI-driven efficiency.

Autodesk's 7% Cut: The Wall Street Journal reports Autodesk is cutting about 1,000 jobs (~7% of its workforce) as part of a global restructuring plan to "strengthen sales and marketing." When a software company cuts 7% of staff to "strengthen" operations, read between the lines: AI is handling tasks that used to require humans.

The Pattern: These aren't struggling companies. Amazon's market cap exceeds $2T. Autodesk is a profitable enterprise software leader. They're cutting because they can, not because they must, and AI is the enabler.

😬 The Entry-Level Freeze

The layoffs are visible. The hiring freeze is quieter, and potentially more damaging to the next generation of workers.

The Financial Times published an analysis of the entry-level hiring squeeze in marketing, communications, and customer service. These are the roles where executives increasingly argue AI can absorb much of the workload.

The Math: Entry-level jobs exist partly to train the next generation of managers. If AI handles the grunt work, companies don't need as many junior employees. But if you don't hire junior employees, where do senior employees come from in 10 years?

The Executive-Employee Gap: A Wall Street Journal survey reveals a stark perception divide:

• Over 40% of executives said AI saved them 8+ hours per week

• Two-thirds of non-management staff said they saved less than 2 hours

The people making headcount decisions are experiencing AI as a massive productivity boost. The people being cut are barely seeing benefits. That gap explains a lot about why the layoffs keep coming.

Translation: The AI productivity story has two versions. Executives see efficiency gains and cut headcount accordingly. Entry-level workers see the ladder getting pulled up behind them. Both are true, and that's the problem.

⚡ Funding Round Quick Hits

🧠 Inferact Raised $150M Seed at $800M: Founded by the creators of vLLM to support and commercialize the open-source inference engine. Led by a16z. When your seed round is $150M, the "seed" label is doing a lot of work.

🎙️ LiveKit Raised $100M at $1B: The company offers tools for building voice, video, and physical AI models, including for ChatGPT's voice mode. Series C hits unicorn status.

💡 Neurophos Raised $110M: Austin-based startup developing a photon-based "Optical Processing Unit" to replace GPUs in AI training. Led by Bill Gates' Gates Frontier. The GPU alternative race is heating up.

☁️ Railway Raised $100M Series B: Cloud platform for software deployment, led by TQ Ventures. Claims $10M+ ARR. Total funding now $130M+.

🔗 Superstate Raised $82.5M Series B: Expanding from tokenized Treasury offerings into SEC-registered equities on Ethereum and Solana. TradFi meets crypto infrastructure.

🔐 Claroty Raised $150M Series F at $3B: Israeli startup making security tech for hospitals, manufacturing, and industrial systems. Critical infrastructure security is having a moment.

🏢 Cambio Raised $18M Series A at $100M: AI-powered commercial real estate software for institutional investors. Led by Maverick Ventures.

💸 M&A & IPO Quick Hits

🍽️ Yelp Acquires Hatch for $300M: Hatch charges a monthly fee for AI-powered agents that respond to customer inquiries and make appointments. Yelp's play to own local business AI.

🏦 Deutsche Boerse Acquires Allfunds for $6.2B: German exchange operator buys European fund distribution platform in cash and stock. Financial infrastructure consolidation continues.

⚖️ Epic v. Google: Secret $800M Deal Revealed: Court filings reveal Epic and Google had a secret deal involving Unreal Engine, Fortnite, and Android, with Epic spending $800M over six years on Google services. So much for the antitrust crusade.

🔒 US House Panel Okays Nvidia Blackwell China Ban: Bipartisan bill pushes congressional oversight of AI chip exports and bans Nvidia Blackwell sales to China for at least two years. The chip war escalates.

🌟 Editor's Note

At Startup Intros, our mission is to bring the latest founder-investor news straight to your inbox, keeping you ahead in the fast-paced world of Silicon Valley.

💭 Parting Thoughts: The Year Starts With a Warning

This week tells a story about what happens when growth slows, and efficiency becomes the priority.

Capital One didn't acquire Brex because fintech is hot; they acquired it because building the same infrastructure in-house would take too long. Amazon and Autodesk aren't cutting jobs because they're struggling; they're cutting because AI makes smaller teams viable. And Big Tech isn't spending $109M on lobbying because they're scared of regulation; they're spending it because regulation is the only thing that could slow them down.

The common thread? The growth-at-all-costs era is over. Now it's about consolidation, efficiency, and buying time. The companies that raised at peak valuations are getting acquired at discounts. The jobs that seemed safe are getting "optimized." And the policy debates that seemed theoretical are now worth nine figures in lobbying spend.

Forward to a friend or hit reply to let me know what you're seeing in your world.

Till next time!

P.S. Raised this week, and we missed you? Want to be featured? Have tips or funding questions? Reply or DM us as we’re here to help.

Startup Intros Weekly Download: Your trusted source for founder-investor insights, delivered with clarity and focus.