Today's headline: AI megarounds hit spend software, databases, and defense. Welcome to Startup Fundraising. American Banker, with Melinda Huspeni:
Ramp has gained another $750 million in its latest fundraise. The New York-based corporate spend management platform announced on Thursday that it is now valued at $44 billion, a 38% increase from its $32 billion valuation seven months ago and a 175% increase from its$16 billion valuation this time last year.
Ramp's Series F — $750 million, $44 billion post-money. That's a 38% jump in seven months, and 175% from the $16 billion mark a year ago. American Banker's carrying it, and they put the AI spend-management product right in the headline. Same three-institution co-lead setup we flagged on the fifth — ICONIQ, GIC, Ontario Teachers'. Confirmed, double-sourced, and still no one is anchoring the price with a single name. Here's my week in one round. Thursday I said the spend platform earns its number, but the AI token-cost add-on is the piece I'd want revenue on before I cheer. Today, it's the headline product the $44 billion is riding on. And the round closed anyway. So the market priced the exact thing I was skeptical of, didn't blink, and now PitchBook's analyst is saying token spend showing up in budget lines makes it 'more valuable.' Show me the budget lines. GIC's interesting here too — same sovereign fund anchoring Supabase this week. Two of the largest rounds in one week, with one institution underneath both. Unite.AI, with Antoine Tardif:
Supabase has raised a$500 million Series F round at a $10.5 billion post-money valuation, marking one of the clearest signs yet that the AI coding boom is changing not just how software is written, but what infrastructure developers choose to build on.
Yesterday we flagged that Supabase had no lead named. Today Unite.AI fills it in — GIC led the $500 million round at a $10.5 billion post-money valuation, with Accel, YC, Craft, Felicis, Peak XV, and Coatue along for the ride. So, question answered. And here's the line that should make you sit up: GIC also co-led Ramp this week. One sovereign fund anchoring two rounds north of $500 million in the same five-day stretch. Say that out loud. Singapore's sovereign fund is deciding the backend layer is where the AI-coding money pools. Bold call on Postgres, I guess. Look, the Firebase-alternative-on-Postgres bet is real — there's a real product there, not just a slide. But $10.5 billion prices in every vibe-coded app paying Supabase rent forever. What's the churn on a developer who spun up a database last Tuesday and bounces by Friday? From Morningstar:
Allen Control Systems(ACS), a leader in autonomous precision robotics, today announced it has raised a $200 million Series B, valuing the company at $2.2 billion post-money. The funding will be used to scale manufacturing and accelerate deployment of its autonomous weapon station, Bullfrog, for the United States and allied militaries, as well as advance development of new product lines.
Allen Control Systems — $200 million Series B, $2.2 billion post-money. And here's the part that matters: Smash Capital led it. The Morningstar wire actually names the lead. Which makes it stand out, because all week I've been pulling mega-rounds with no lead named out of this rundown. This one gives me a name and the syndicate — Craft, Rally, Inspired. So, credit where it's due. Okay, but $2.2 billion at Series B for a company whose product bolts AI onto an M240 machine gun? I ran this same math on Mach Industries. 'Selected by U.S. and Allied militaries' is a contract signal, sure — but 'selected by' isn't a multi-year program of record. Bullfrog's already deployed with the Army and Navy, 100 percent success rate in the readiness experiment — that's real. But DoD procurement runs on a five-to-seven-year calendar, and that valuation prices in the program before the budget's even appropriated. The wire doesn't give a prior-round valuation either, so I can't tell you the step-up. Big number, no anchor to measure it against — that's the recurring gap, even on a deal that disclosed its lead. This one's from Beinsure:
Pace, an AI operations partner for insurers, has raised $46 mn in Series B funding to expand its agentic workforce across insurance operations. Thrive Capital and Sequoia Capital co-led the round, with Emergence Capital and Pruven Capital also participating.
Pace, $46 million, Thrive and Sequoia co-leading — and the pitch is an 'agentic workforce' for insurance ops. That phrase makes me nervous right away, because it's the same kind of hand-wave as 'AI spend management' yesterday. But — credit to them — there's actual evidence under it. 250,000 workflows completed, claim cycle times cut 30% with Ryze, customers like Prudential and WTW. That's the unsexy plumbing I actually like. And the lead structure is clean for once — Thrive and Sequoia both named, both co-leading. No mystery anchor. After this week, naming your leads counts as a feature. Watch the Thrive angle, though — it's co-leading here while it led Helion yesterday. Same top-tier fund showing up as early backer and growth-round price-setter across the slate. That's the Benchmark conflict question, live. My question on Pace is liability. When the agent makes a bad call on a claim, who owns it — Pace or the insurer? 'Autonomously completed 250,000 workflows' sounds great until one of them denies a claim it shouldn't have. From Fund Momentum:
Merantix Capital, the investing arm of Berlin's Merantix AI group, has closed a €103 million fund to back early-stage, AI-native teams across Europe. It is the firm's largest vehicle to date — roughly three times the size of its first fund, a venture-studio-only vehicle publicly reported at €30-35 million — and it is structured unusually: half the capital funds founders Merantix helps build from the pre-idea stage, and half goes into direct pre-seed and seed deals.
Merantix Capital, out of Berlin, closed €103 million for early-stage European AI — roughly three times their first fund, which ran €30 to €35 million as a studio-only vehicle. The structure's what I want to pull apart — half the money builds founders from pre-idea inside their own campus, and half goes to outside pre-seed and seed. It's a studio and a seed fund under the same roof. And I actually like the thesis — logistics, manufacturing, energy, robotics. They're betting Europe's edge is the physical economy it already runs, not another chatbot. That's a real bet. Tripling your fund size is the move everyone's making this week. Benchmark did it, now Merantix — same stated reason, expand to compete. So are forty companies at pre-idea conviction, or just more shots? Forty targets off €103 million, with half of it pre-idea — that's thin checks per name. They're betting the campus does the de-risking more than the capital does. Got a fundraising question, a story idea, or a correction for us? Send it our way at startupfundraising at lantern podcasts dot com. We read every note, and your feedback helps shape future briefings.
You’ll find links to all of today’s stories in the show notes, so if something sparked your interest, that’s the place to dig in a little more.
That’s Startup Fundraising for today. Thanks for listening, and enjoy the rest of your Saturday. This is a Lantern Podcast.