TD-01Capability
Unified corpus
Sporting, technical and financial regulations, Technical Directives and stewards' decisions, tagged with context and outcome.
DOC 001 · Regulatory Intelligence for F1
From idea to calculated risk.
Before a part is built, APEX estimates how the FIA would read it: cites the exact precedent and Directive, and drafts the argument — weeks before the problem exists. The same engine builds the defence on race day, when protest or defend is decided in minutes.
DOC 12 · APEX POINT
73%
Estimated probability of sanction
4 precedents cited
DOC 60 — Impeding — Car 10 vs Car 18Reprimand — Car 44 — Yellow flag infringement5s Penalty — Car 27 — Left track and gained advantageWarning — Car 30 — Abnormal change of direction3-grid penalty — Car 10 — Q1 impedingWarning — Car 23 / Car 41 — SC2-SC1 time exceedanceDOC 60 — Impeding — Car 10 vs Car 18Reprimand — Car 44 — Yellow flag infringement5s Penalty — Car 27 — Left track and gained advantageWarning — Car 30 — Abnormal change of direction3-grid penalty — Car 10 — Q1 impedingWarning — Car 23 / Car 41 — SC2-SC1 time exceedance
DOC 002The problem
Risk is decided on the design desk, not on Sunday.
Risk
Blind design decision
Engineering commits a geometry without systematically cross-checking it against years of Technical Directives and precedents. It's decided on the reading of two or three people.
Risk
Permanent change
A Technical Directive can shift the picture from one Grand Prix to the next, and the impact is assessed by hand every time.
Risk
On race day, even less margin
If the risk wasn't calculated earlier, protesting, defending or taking the risk is decided in minutes, against the clock.
DOC 003The solution
Four capabilities on a single corpus.
TD-02Capability
Predictive engine
Probability of penalty, protest or successful defence, calibrated against cases already resolved. Used on the design desk, weeks before a race, and also on race day.
TD-03Capability
Defence simulator
Anticipated argument before the stewards, with cited and verifiable precedents, before you file the case.
TD-04Capability
Mutation alerts
Automatic notice when a new Technical Directive shifts the weight of a live precedent.
DOC 003.5Illustrative example · Not a prediction
How APEX structures a real case.
This case has already been decided by the stewards — we use it to show the analysis format, not as a forecast. The first case predicted before the actual ruling starts with the Belgian Grand Prix, on 17 July.
FIA · Steward decision
- Source
- DOC 60 — FIA, 2026 British Grand Prix
- Fact
- In Q1, Car 10 was on a slow in-lap on the racing line, forcing Car 18 — on a timed lap — off line to pass.
- Applicable article
- Unnecessary impeding of a car on a fast lap.
- Mitigating factor considered
- Radio failure on Car 10, so the driver was not warned of the approach.
- Actual outcome
- Standard 3-place grid penalty.
APEX · Análisis
This is how APEX would structure the case — with no invented probability, because the outcome is already known and showing a % would be misleading.
- Applicable precedent
- Impeding in qualifying — typical sanction: 3 grid places, downgraded to a reprimand when a technical mitigation exists.
- Weight of the mitigation
- Radio failure documented via telemetry — strong mitigation, does not exonerate but dampens escalation.
- Argument
- Acknowledge the impeding; argue communication failure outside the driver's control; request a reprimand instead of a grid penalty.
In the actual case, this argument was not enough to avoid the standard penalty — the exonerating telemetry weighs as a mitigating factor, but does not exonerate.
Live track record
Next: Belgium, 17–19 Jul
Reserved for the first prediction issued before the stewards' ruling. Will replace this illustration when the weekend closes.
DOC 004Why it isn't copied
The moat is the data, not the software.
§
The data
Structuring decades of precedents is work no in-house team will take on.
§
Inverse network effect
If your rival uses it and you don't, you race with asymmetric information.
§
Neutrality
An external vendor sees the aggregate pattern of how the FIA rules across every team.
DOC 005Tiers
Anchored to the value at risk.
Most chosen
Pro
USD 350,000–500,000/ year
Everything in Core + defence simulator + mutation alerts.
Request pilotElite
USD 600,000–900,000/ year
Everything in Pro + trackside support + assigned analyst.
Request pilotDefence module can be activated per Grand Prix, available as an entry point with no annual contract.
DOC 006The proof
Before you sign, we test it against the past.
We take 15 to 20 stewards' cases already resolved. The engine predicts the outcome blind. It's compared against the actual ruling, in the room, before any economic commitment.
- 01Case selection with a documented knowledge cutoff.
- 02Prediction sealed before the ruling is revealed.
- 03Compared against the informal vote in the room.
- 04Only then do we discuss pricing.
Commitment · Calibration
≥ 80%hit rate on blind cases, before we invoice.
If the engine doesn't clear the threshold in your pilot, there is no contract. Calibration is signed off with you — not declared by us.