Top 10 VC Deals This Week
Week of Nov 10–17, 2025
Top 10 VC Deals This Week
Cursor raised $2.3B at $29.3B valuation, Ramp hit $32B managing $10B monthly with AI agents, and defense tech grabbed $510M. This week venture capital shifted from AI experiments to AI infrastructure. Here’s what matters for your strategy.
1. AI Coding Startup Cursor Raises $2.3B at $29B Valuation
November 13 | Four MIT grads just raised one of the largest AI rounds ever. $2.3B Series D at $29.3B valuation—tripling in 6 months.
Not GitHub Copilot bolted onto old tools. Native AI-first IDE.
$100M ARR. Majority of Fortune 500 using it. Enterprise revenue up 100x year-over-year.
Why you care: AI coding tools crossed the chasm. Your dev teams not using them? Competitors are shipping 100x faster than you.
Sources: Business Wire, Bloomberg, Crunchbase
2. Defense Tech Unicorn CHAOS Secures $510M at $4.5B Valuation
November 13 | Just 3 years old. Los Angeles-based CHAOS raised $510M for autonomous defense systems and counter-drone tech.
Software-first command-and-control. Updates as fast as mission requirements change.
Led by Valor Equity Partners.
Why you care: Defense procurement is moving from decade-long cycles to software speed. Governments want AI capabilities now—not in 2035.
Sources: Crunchbase
3. Ramp Hits $32B Valuation as AI Agents Manage $10B+ Monthly
November 17 | $300M round led by Lightspeed. Up from $22.5B valuation in July.
AI agents made 26 million decisions across $10B+ in spend last month alone. Prevented 511,000+ out-of-policy transactions.
50,000 customers. $1B+ annual revenue. Companies using Ramp spend 5% less and grow 12% faster.
Why you care: Autonomous finance is here. AI isn’t flagging problems—it’s preventing them. CFOs: build in-house or adopt platforms where AI is the product?
Sources: Bloomberg, PR Newswire, Crunchbase
4. AI Chip Startup d-Matrix Raises $275M to Challenge Nvidia
November 12 | $275M Series C at $2B valuation. Santa Clara startup targeting AI inference—where models run in production.
In-memory compute architecture. Addresses the “memory wall” bottleneck.
Backed by Microsoft Ventures, Qatar Investment Authority, Temasek.
Why you care: Inference costs exceed training by 10-100x at scale. Specialized chips will fragment Nvidia’s dominance. Your infrastructure strategy needs multi-vendor chip plans.
Sources: Tech Startups, Crunchbase
5. Fireworks AI Secures $250M as Inference Becomes the Battleground
November 13 | $250M Series C at $4B valuation. Redwood City platform processes 10 trillion tokens daily for 10,000+ customers including Uber, Shopify, Cursor.
$280M ARR. Founded by PyTorch team. Multi-model routing and orchestration—no cloud lock-in.
Nvidia and AMD both invested.
Why you care: Inference efficiency determines AI economics. Multi-model approach beats vendor lock-in. Evaluate whether hyperscaler AI services actually deliver value.
Sources: Fireworks AI, Business Wire, PYMNTS
6. Commercial EV Maker Harbinger Lands $160M With FedEx as Lead Investor
November 13 | $160M Series C co-led by FedEx, Capricorn, THOR Industries.
Not consumer EVs. Purpose-built for fleets. FedEx ordered 53 vehicles for 2025 delivery.
Why you care: Fortune 50 fleets are committing capital AND vehicle orders. Medium-duty EVs hit economic inflection. Electrification moves from pilots to fleet-wide rollouts in 2025-2027.
Sources: Crunchbase
7. Sensor Startup Teradar Emerges With $150M for Autonomous Vehicles
November 12 | $150M Series B for MIT spinout. Terahertz sensors combine radar + lidar performance at mid-market price.
Already working with 5 top automakers. Contracts expected for 2028-model vehicles.
Dual-use: automotive and defense. Lockheed Martin Ventures invested.
Why you care: Sensor costs bottleneck ADAS adoption. This could enable affordable autonomy for mainstream vehicles—not just luxury models.
Sources: TechCrunch, Crunchbase
8. Marketing AI Startup Alembic Raises $145M on Causal AI Breakthrough
November 13 | $145M at $645M valuation—15x increase from last round.
Not correlation analytics. Causal AI identifies what actually drives revenue vs. what coincides with it.
Jensen Huang read coverage and initiated Nvidia partnership directly. Now deploying an NVL72 superPOD—one of the fastest private supercomputers.
Why you care: CMOs finally get AI that proves which spending actually works—not just dashboards showing what happened. Marketing optimization moves from guesswork to science.
Sources: VentureBeat, Crunchbase
9. Nuclear Startup Valar Atomics Raises $130M to Power AI Data Centers
November 10 | $130M Series A led by Snowpoint Ventures. One of the largest nuclear Series A rounds ever.
Building reactor “gigasites” for AI data centers. Palmer Luckey (Oculus founder) invested.
Why you care: AI compute is power-constrained, not economics-constrained. Data center locations now determined by energy availability. Partner with utilities NOW for adequate supply.
Sources: Tech Startups
10. AI Agent Platform Wonderful Races to $700M Valuation in 4 Months
November 11 | $100M Series A at $700M valuation. Amsterdam-based (R&D in Israel). Fastest path to unicorn-track in 2025.
Multilingual AI agents with cultural fluency. Managing tens of thousands of daily requests at 80% resolution rate.
Expects $10M ARR by year-end.
Why you care: Customer-facing AI agents work TODAY with enterprise-grade results. Global support without local teams. Deploy fast or competitors will capture the cost advantages.
Sources: TechCrunch, Tech Startups
Bottom Line
AI infrastructure captured 67% of funding this week—$3.7B+ in “picks and shovels.” Investors bet on infrastructure over applications, mirroring cloud computing’s value capture.
Defense tech grabbed $700M globally. Software-first military systems disrupt traditional contractors operating at 1/10th the speed.
Vertical AI hit product-market fit. Customer service, legal, sales, and finance AI showing measurable ROI today—not theoretical future value.
Energy infrastructure emerged as AI bottleneck. Power availability now constrains AI deployment more than chips or algorithms.
The inference cost crisis is real. Multiple $250M+ rounds targeting inference efficiency. Production AI costs exceed training by 10-100x.