One system of record for athlete operations
Athletic departments run recruiting, contracts, budgets and compliance in separate places, spreadsheets, inboxes, and whatever the last administrator set up. Sherpa replaces that with one platform. We built major modules of it, and the language-model layer that reads the documents flowing through it.
Four domains, one source of truth
Scouting, contracts, finance and compliance are usually four separate systems that disagree with each other. Putting them in one platform is less a UI problem than a data-modelling one: a recruit, a signed NIL deal, a payment and an audit record are all views of the same athlete, and the moment they drift apart the compliance function stops working.
Scouting & recruiting
Prospect records, evaluation notes, communication and offer workflows held in one pipeline instead of scattered across staff inboxes.
NIL & pro contracts
Deal lifecycle from draft through signature, with versioning, the part where a lost document becomes a regulatory problem rather than an inconvenience.
Financial operations
Budget planning, athlete payments and spend tracking tied back to the deals and people they belong to.
Compliance
Rule monitoring and audit trails generated as a by-product of normal use, rather than assembled by hand before a review.
The documents are the hard part
Athlete operations runs on documents, contracts, offer sheets, evaluation write-ups, policy PDFs. They arrive as prose, and everything downstream needs structure. That gap is where the language-model work sits.
Extraction
Pulling the terms that matter out of contracts and agreements and turning them into records the platform can enforce rules against, rather than a file somebody has to open and read.
Summarisation
Condensing long records and histories into the version a coach or compliance officer can act on, without losing the detail underneath it.
Conversational search
Asking questions of the data in plain language, so finding an answer does not require knowing which of four modules stores it.
Grounded in their data
Answers trace back to the underlying records. In a compliance context, an output nobody can verify is worse than no output at all.
Reviews
