Why Handoff

Data is the moat.

Schemes get copied, staff get hired away, and every program can buy the same software. What does not transfer is your own accumulated record of what happened and what you decided. That is the durable advantage, and in most departments it is unusable.

Why it is not working yet

The data already exists. It is just messy, scattered across systems that were never built to talk to each other, and not in a state where anything can reach it on demand. Recruiting sits in one product, performance in another, compliance in a third under different rules, and the connective work happens in a spreadsheet somebody maintains by hand.

Scattered

Six products that do not speak, so any question crossing two of them becomes an export.

Messy

The same athlete under three spellings, the same event recorded three ways, no agreed key.

Not on demand

Answers take days because a person has to assemble them, so most questions are never asked.

Frozen in time

An export is a copy that starts going stale immediately. Decisions get made against copies.

What it takes

A technological edge needs a clear, canonical picture of your program that evolves in lock step with it. Not a warehouse somebody loads once a quarter, and not a dashboard sitting on top of the same disconnected sources. One record of what is true, updated as the season moves, that models and agents can both read.

That is the work Handoff does. We ingest what your department already produces, resolve it to one schema with owner, state and evidence attached, and position it so it is useful to the people and the systems that need it.

Two paths in, one record out

Data reaches the record two ways, and they are kept separate on purpose. Shared league data is ours to run: we maintain the adapters, the ingestion and the canonical form, so no school pays the cost of keeping the ecosystem current. Everything your department produces arrives through connectors you install and control, and lands in a store scoped to you alone.

Global sources, managed by us

Play by play, schedules, recruiting, the portal and downstream outcomes come in through source adapters we maintain and platform ingestion we run. They resolve to global canonical data: athletes and their external IDs, league plays, recruiting history, portal history and the inputs behind national metrics.

Org sources, managed by you

Teamworks, Catapult, ARMS, your private PFF, internal databases and plain uploads come in through connector definitions you install per system. They resolve to org canonical data: roster, evaluations, NIL, workload, wellness and internal scouting. It is yours, and it stays in your scope.

The two meet in one computation layer. Global metrics are computed once across the whole population; org metrics are computed against your records alone; and the numbers worth acting on come from reading one against the other. Models sit on top of that, and the agents and workflows your staff uses sit on top of the models, never on a raw feed.

What that unlocks

Agents that reason across recruiting

A specialist that reads your board, your evaluations and your contact history together, and coordinates a recruiting visit end to end rather than answering a question about one.

Models for the transfer portal

Custom scoring your staff builds and can inspect, pointed at finding undervalued players your scheme specifically wants.

Performance

Load, availability and production in one place, so the training question and the depth chart question are asked of the same record.

Wellness

Availability and workload held under their own permissions, reaching only the roles cleared for them.

Compliance

Eligibility, contact logs and approvals recorded as they happen, with what the approver saw at the time.

Operations

Travel, lodging, scheduling and staff assignments as records rather than as a thread somebody has to remember.

Scoped from the whole population down to you

The floor is every school, every player, every sport. Each tier above it narrows, and your own records are the tier that makes the answer yours rather than generic. That is why a school specific number still carries the weight of the whole population behind it.

AgentsClaude, ChatGPT, or your ownRuntimesModels, apps, API, any toolIntegrationsRead and write, both waysTeam modelYour program's viewGlobal modelEvery school, every playerWidest at the base. Everything above it is a narrowing.

It compounds, and it stays yours

The store is worth more every season it runs, because the advantage is the accumulation. Three years of your own evaluations joined to what actually happened is something no competitor can buy and no departing staff member can take with them.

It is yours. Your records, your models and your program's work stay your property, whether or not you stay a customer. A moat you rent is not a moat.

See the design partner programRequest access

Or read analytics for what we compute on top of it.

Back to the overview