One record, not a dozen versions of the same founder

Every tool keeps its own version of the founder

The founder who applied, the founder who was scored, the founder an advisor met and the founder in the board report are usually four different rows in four different tools. Nobody planned it that way. Each tool was bought for one job, and each one started its own copy of the founder.

Every disagreement between those copies costs somebody an afternoon. A program manager decides which revenue figure is current. Someone rebuilds the funder report from exports that don’t share a common ID. The founder fills in a fourth intake form that asks what the first three asked. And when partners in a region pool their numbers, one founder gets counted by every organization they touched, with no way to say how far the total is inflated.

Most sectors settled this long ago. Hospitals keep one authoritative record per patient, sales teams one per customer, and HR one per employee. Every other tool in those sectors reads from that record, and when two tools disagree, the record wins. Founders are the one major economic role that never got an equivalent.

We built the platform around one record per founder

So we made the founder record the product and treated each module as a way of working on it. An application creates the record. Reviewers score against it in Evaluation and Scoring. Advisors log their sessions on it. When a founder redeems an offer, Redemption Tracking logs the redemption on the same record. Reporting Dashboards read from it. Nobody re-enters the founder at any step, so there’s no second copy to drift away from the first.

Different seats see different parts of it. Permissions, set by your team, decide what an advisor, a sponsor or a staff member can open, and nobody keeps a private copy to reconcile later.

We also chose not to model the founder as a deal. A CRM is organized around something that closes. A founder carries on across companies, programs and years, and the record has to follow the person through all of them.

The record keeps events, and reports are views of them

The design choice that matters most is a quiet one. The record stores what happened: this founder, this session, this date, this advisor, this offer. Interpretation comes later, when someone asks a question.

That’s how one program can answer several funders without arguing with itself. Your federal grant counts a business one way, your foundation sponsor another, your board a third. Each of those definitions is a different view of the same events, so the three reports can count differently and still agree on the facts underneath them. If a definition changes next year, the events it applies to are still there.

It matters for AI as well. Nova, the platform’s AI, answers founders and program teams from this record and shows which entries it read. An assistant working from four disagreeing rows can only guess which one is current. Working from one record, it can answer and point to its sources.

The founder keeps the record when the cohort ends

A founder’s work runs longer than any program. Most of the outcomes a program would like to claim, the large round, the first real hires, the exit, form two to seven years after the cohort ends. The hard moments come late too, like a co-founder leaving eighteen months after demo day. By then the program has usually moved on to its next cohort.

So the record follows the founder from program to program. When a cohort ends, the founder keeps their record and keeps using the platform free. Your program keeps a reachable alum instead of a stale spreadsheet row. If that founder comes back during a crisis, or later joins another program that runs on Startup Science, they start from what they’ve already built.

If you want to see how far your own versions have drifted, take one alum from two years ago and look them up in every system your team uses. Count how many of the answers still agree.