Why it exists
Engineers leave most sessions with more data than time. The hard part is deciding what to change next, and knowing why. Racing is built around that decision: compare fairly, keep every explanation tied to its evidence and write one test that can tell the explanations apart.
What it does today
Racing is in private early access. Every workspace uses one synthetic Jerez session with a fictional GT4 car, #27, so you can judge the workflow without sharing any data.
- A comparison of two runs, with lap exclusion and run medians that update.
- Evidence inspection through the sector 2 corner window: speed, throttle and brake.
- AI driving and setup hypotheses from the OpenAI API, each with its cited evidence and the measurements that are missing.
- Context correction and a revised analysis that keeps its source revision.
- One editable, saved next-test plan.
- A private workspace for each invitee.
What it does not do yet
It does not import your telemetry, and it makes no lap-time claims. Import and a per-team model are planned with design partners for the 2027 season.
Design-partner roadmap, 2027 season
Racing is being shaped with a small number of design-partner teams for the 2027 season. None of this is available today.
- MoTeC and AiM CSV import.
- A per-team isolated model trained only on that team's data, with no cross-team training.
- Team data stays the team's, with export and deletion on exit. Data ownership describes how data is handled today.
Racing and Apex Scout
Apex Scout at apexscout.app helps track-day drivers across Europe find events and prepare for them. Racing is a separate product for race engineers, with its own accounts, database and API. The two share a founder and a design language, not your data.
Who builds it
Racing is built by João Paulo Santos in Leiria, Portugal. Every contact message reaches him directly.