The best one this quarter is somewhere in your inbox, in a deck nobody opened. Here, every application gets read properly — structured on the way in, checked against public sources, and scored against your thesis. You decide from a memo, not from a PDF at midnight.
No card required · Live application link in under a minute
# What a founder pastes into Claude, Cursor, or ChatGPT Read https://yourfund.pitchslay.com/skill.md # That's the entire integration. Nothing to install. ✓ Reads your live schema ✓ Interviews the founder from their deck ✓ Submits a clean, validated application
Then an analyst spends their week turning PDFs back into fields nobody verified.
Put one line on your Apply page. Your schema, your questions, your brand — and a form for founders who prefer one.
Their assistant reads your live schema, interviews them from their own material, and submits something clean and complete.
Research runs, signals get verified, the deal is scored against your thesis, and you take a decision in one click.
The same engine, configured for how you actually work.
Rolling deal flow scored against your thesis. Off-thesis deals never reach your inbox, and your override rate against the engine is measurable from day one.
Cohort programs with real windows and capacity. Each program gets its own schema, its own thesis, and its own selection bar — and structured data your scheme reporting can actually use.
Batch season without the batch panic. Applications arrive in the last 48 hours; the pipeline queues, drains, and ranks the whole cohort against itself.
You do not have an analyst. The memo is the analyst — verified signals, contradictions surfaced, and a recommendation you can disagree with on the record.
Several funds, several theses, one platform. Separate programs, separate teams, separate decision rules, one billing relationship.
Scouts submit on behalf of founders and the provenance is recorded. Route matched deals to the investors in your network by stage, sector and cheque size.
Nothing in the console is a stored magic value. Score, conviction and thesis match are all projected from a table of signals — each one categorised, weighted, and sourced.
Strengths, risks, contradictions and gaps are all derived from this table.
Founders apply through Claude, Cursor, ChatGPT or any assistant that can read a URL. No install, no OAuth, no account for the founder.
A locked core every investor asks for, plus unlimited fields of your own. Relabel, reorder, make things required. Every artefact regenerates automatically.
Stage, sector, cheque size and geography. Out-of-thesis deals route themselves into their own bucket instead of your reading list.
Set the thresholds for take-meeting, follow and pass. A live distribution shows how your real applications fall before you commit.
Rolling intake or dated cohort windows with capacity. The skill file states the live window, so an agent never creates an orphan draft.
Owner, reviewer, analyst, read-only. Every field change, status move and decision is recorded with who and when.
Application volume and pipeline cost tracked per month, per program, so you always know your cost per deal reviewed.
Signed webhooks with a delivery log, rotatable API keys, and CRM connections for Affinity, Airtable and HubSpot.
Accent colour, custom domain, and application pages that look like yours rather than ours.
Every plan includes a 14-day trial. No card to start.
One program, a small team, and enough volume for a first cohort.
Several programs running at once, with the whole investment team on it.
Batch-season volume, unlimited programs, and priority research capacity.
Neither. The agent route works with whatever assistant the founder already uses, and there is always a plain web form as a fallback. Identity is established by a six-digit email code, not a login.
They fill in a form generated from the same schema, and it goes through the same pipeline. Nothing about the review differs.
Yes. There is a locked core covering what every investor asks, and everything else is yours to define. You can relabel core fields, reorder them, add help text, and make optional ones required.
Email verification, rate limits, disposable-domain blocking, and per-domain caps. Beyond that, applications with no resolvable company and no verifiable founder footprint are flagged for you — never silently rejected.
You do. Applications submitted to your programs are yours, exportable, and never shared with another fund on the platform without your explicit opt-in.
No, and deliberately so. We are infrastructure. We do not run a fund, we do not take carry, and we are never competing with you for a deal you sourced.
Set up your program, publish the link, and see what comes back. Fourteen days, no card.