Don't miss the next
Shayne Coplan
With Claude, anyone can spin up a product, a deck, and basic usage in a weekend. Your inbox is going to be 100x in 6 months. The next breakout founder is in there. So is a tidal wave of noise that looks identical to it.
Let your AI agent do the 80/20 filtering. Built on your real investment thesis, not a generic LLM with a generic prompt. So the breakout founder gets to the top of your queue instead of buried under 1,000 vibe-coded landing pages.
Free for the first 10 investors. We onboard you personally.
Step 1 — find yourself
We've probably already drafted your profile.
Search your name or your fund. If we have you, click through to claim it. A human will verify within 24 hours.
We've already indexed hundreds of crypto VCs. Yours might be here.
The problem
The bar to ship a "real-looking" startup
just collapsed.
A working product, a polished deck, a landing page with three logos and some traction metrics, all of it now takes a weekend with Claude. Most of what hits your inbox in 2026 will look indistinguishable from the real thing. The signal-to-noise ratio doesn't degrade gracefully. It collapses.
decks per month at a typical seed fund today
in 6 months once "build a startup" costs a weekend
missed Polymarket is the cost of not filtering well
How we build your agent
Your agent is only as good as your profile.
So we make the profile real.
Most "investor matching" tools rely on a one-line bio and your fund's sector tags. That's why they're useless. Our agent is built on three layers of context, in this order.
We seed it from public data
Your last 50 investments, your portfolio sectors and stages, your writing, your podcast appearances, your tweets, the rounds you led vs. followed. RaiseKit has been collecting this for months across hundreds of crypto VCs. We start your profile already 60% full.
No questionnaire to fill out before you've seen value. Sign up, see the draft, then correct what's wrong.
You fill the gaps and correct the wrong stuff
The public data has blind spots. You stopped doing seed last year. You moved off consumer crypto. The thesis the agent inferred is from 2023 you, not 2026 you. You edit, in plain English. The agent rebuilds.
Profile gets richer every time you correct a filter decision. The bad calls become training signal.
We do a 1-hour call to capture how you actually decide
This is the part nobody else does. We sit with you for an hour and ask the questions no questionnaire can capture. Why did you pass on X? What made Y a yes after a 20-minute call? What's the founder profile that you'd back even when the deck is mediocre? What patterns do you see that your partners don't?
Recorded, transcribed, summarised into the agent's prompt. This is what makes your agent yours instead of a generic VC agent.
What lands in your inbox
A daily queue. Sorted. Scored. Reasoned.
✓ The 5 that fit
Each pitch comes with a one-paragraph why-this-fits, scored against your thesis, your stage, your check size, and the patterns from your call. If the founder matches a "yes even with a mediocre deck" pattern you mentioned, that's flagged.
— The 195 that don't
One-line auto-passes with the reason. Wrong stage. Wrong sector. Wrong geography. Wrong check size. You can override any of them with one click, and the override teaches the agent. Founders get a polite no with the actual reason.
~ The 10 borderline
The interesting cases. Adjacent to your thesis. Founder you'd want to meet even if the company isn't a fit. A new sector that's starting to look like an old sector you already back. These are the ones to look at first.
⚡ The pattern report
Weekly: what showed up in your inbox this week, what your peers are seeing, where your filter is diverging from the market. The kind of thing your partners care about at the Monday meeting.
Why this works now
Generic LLMs can't do this. A model with your context can.
The reason every "AI for VCs" demo feels off is that they're running on a generic model with a generic prompt. That model doesn't know you passed on Solana infra companies in 2023 and changed your mind in 2025. It doesn't know your fund leads and writes 2M cheques but only at pre-seed. It doesn't know your partner hates founder-as-CEO when the co-founder is the actual product person.
We give the model that context. Then it can do the actual job. That's the whole pitch.
The honest part
Where we are, what's next
We've already built the data layer. RaiseKit indexes hundreds of crypto VCs, their writing, investments, theses, and check patterns. The agent reads from this.
Onboarding the first 10 investors manually. 1-hour call, agent built, pitch ingestion wired up to your inbox or a forward address. Free during the design partner phase.
Pricing model lands once we know what the time savings are actually worth. Likely per-seat at a fund, with a discount if your whole partnership uses it.
Want one of the first 10 spots?
Two ways in. Claim your profile and get an instant dashboard, or DM me directly.
No form. No waitlist email. Claim your profile or just DM.