Get in touch
← All work

Sports data analytics

From sensor data to race-winning decisions

We turn raw, multi-channel sensor streams from high-performance sailing and foiling into clear, decision-ready insight — and into the tools that put that insight in the hands of athletes and engineers.

Confidential: SailGP engagement data is private. Any charts on this page use generalised or synthetic data and illustrate method only — no client data is reproduced.

The challenge

High-performance sailing and foiling generate large volumes of multi-channel telemetry — boat speed, load, attitude, control inputs, environmental conditions. The raw data is rich but unusable as-is: the value is locked in the relationships between channels and in the handful of moments that actually decide a race.

What we did

  • Built data pipelines to ingest, clean, synchronise and process raw sensor data into analysis-ready datasets.
  • Performed insight extraction — deriving the performance metrics and patterns that matter to athletes and engineers.
  • For SailGP, developed an interactive website with a React-based AI agent that answers concrete questions about the discipline by querying a curated knowledge base (a “second brain”).

Approach

1 — Data pipeline

Placeholder: ingestion, cleaning, time-alignment of channels, and the processing steps that produce a trustworthy dataset.

Time-series
PlannedSynchronised multi-channel sensor traces for a single run, annotated with the key events.

2 — Insight extraction

Placeholder: the derived metrics and the relationships that turned out to drive performance.

Performance polar
PlannedSpeed vs. angle/condition, showing where performance is won and lost.

3 — Interactive tools & AI agent

Placeholder: how the interactive tool and the question-answering agent let the team self-serve answers instead of waiting on analysis.

Product screenshot
PlannedThe interactive dashboard and the AI agent answering a domain question.

Outcome

Placeholder: the decisions this enabled and the difference it made.