Racing Data

App Home Page

Inspiration

Motorsport is full of data, but most drivers never get to see it in a way they can understand. They know when a lap feels good or bad — but not why. Small mistakes, braking too early, or losing momentum in a sector are hard to detect without proper tools.

Racing Data was born from that gap. My goal was to create an intuitive, AI-powered platform capable of turning raw TRD telemetry into actionable insights that any driver can instantly understand.

What we built

Racing Data is a web platform that analyzes a driver’s performance using official TRD datasets. It transforms speed, throttle, braking, and timing data into:

AI-generated performance score (0–100)
AI race summary, giving a general snapshot of the driver’s race
Position table, showing the selected driver's finishing place in context
Telemetry visualization, compared against the race-average driver profile
Sector analysis, comparing best sectors to sector records and analyzing average telemetry within each sector

The platform is fast, simple, and designed for drivers who want clarity — not spreadsheets.

How it works

1. Home → Race selection

The user starts on a minimal homepage. They select a race (for the MVP, two races are loaded, but the system is fully scalable to any season).

2. Driver selection

After choosing a race, the user selects a driver. Only then does the platform begin its analysis — everything is personalized.

3. Dashboard

Once a driver is selected, Racing Data generates:

AI Performance Score: A quick indicator of overall performance.
AI Race Summary: A general narrative describing how the race went for that driver based on their data.
Position Table: Shows 11 key positions from the race, including the driver’s exact finishing position.

4. Telemetry View

Telemetry charts show:

Speed
Throttle pedal position
Brake pressure
And more

All telemetry is calculated relative to the race’s average driver profile, allowing direct comparison between the selected driver and the “average” performer of that race.

5. Sector Analysis

The race is divided into Sector 1, Sector 2, and Sector 3.
For each sector, the platform:

Compares the driver’s best sector to the sector record of the event
Computes average telemetry within the sector’s time window, helping reveal where improvements were possible
Highlights where time was gained or lost

This gives drivers a clear understanding of how they performed, not just what their times were.

Challenges we faced

Data complexity: TRD datasets are large, raw and not directly usable. Processing them into structured telemetry and timing required careful design.
Performance: Visualizing telemetry without slowdowns demanded optimization and caching.
AI integration: Generating meaningful summaries and performance scores required fine-tuning prompt engineering to avoid hallucinations.
Scalability: Although the MVP uses two races, the backend had to be built with full season support in mind.

What we learned

How to transform motorsport telemetry into understandable insights
How to build a clean front-end experience on top of a powerful data pipeline
How to merge AI-generated analysis with technical data in a safe, controlled way
How to design with clarity and simplicity as priorities