Useful intelligence, not demos.
Agents, inference infra, vertical copilots that compound data. Having shipped LLM systems at Google, I can tell demo-ware from product-ware in 10 minutes.
I'm Karl Qu — AI engineer at Google by day, venture scout by night. I back ambitious technical founders across AI, space, and fintech, and connect them with the right checks at seed and Series A.
Three focus areas where I have the deepest networks and highest conviction —
Agents, inference infra, vertical copilots that compound data. Having shipped LLM systems at Google, I can tell demo-ware from product-ware in 10 minutes.
Launch economics, in-space services, Earth observation, and dual-use defense. Patient on physics, impatient on go-to-market and unit economics.
Infra for payments, treasury, and underwriting; vertical fintech where software eats a pre-software P&L. I care about distribution more than features.
I read everything that lands in my inbox. Here's the shortest path to a real conversation and an intro to a fund.
What you're building, for whom, and why now. Link to a live product or a 5-slide deck — whichever shows more.
The single number that proves you're onto something — retention curve, unit economics, waitlist growth, a signed LOI.
If there's fit, we do a no-slides call. I ask mean questions so the real VCs don't have to.
I double-opt-in every intro. You get context; partners get conviction. Expect 2–4 intros within the week.
I currently scout for a handful of pre-seed and seed funds. If your fund is looking for a focused, technical scout in SF with deep coverage of AI, space, and fintech — let's talk.
Currently writing — Space is cheap again · What I got wrong in 2025 · A scout's contract, explained · Fintech's boring middle
Tell me what you're building in a paragraph. If there's a fit with my thesis or network, I'll reply within 72 hours.
Happy to chat about scout programs, co-investment, or just compare notes on what we're each seeing in AI, space, and fintech.