200M
Performance intelligence — Built for the elite track

Every watt,
every lap,
accounted for. A governed intelligence layer between the bike and the program. Capture telemetry. Reason on it. Decide faster.

Built for National federations, performance directors, and the athletes they back.
Discipline Track sprint · Team pursuit · Mass start · Keirin · Kilo
Status Closed pilot. Federation engagement by invitation.
Geography North America · Caribbean · UK pipeline · Pacific 2026
LAP 14.236s · 71.4 km/h · 1284 W · NJP/01 200M 10.412s · 69.2 km/h · STD START · KBR/02 TPP LAP 17.81s · 56.7 km/h · 412 W AVG · PURSUIT/A FLY 200 9.872s · 72.9 km/h · 1456 W PEAK · NCP/03 KILO 1:00.412 · 59.6 km/h AVG · 647 W · MKW/04 LAP 14.236s · 71.4 km/h · 1284 W · NJP/01 200M 10.412s · 69.2 km/h · STD START · KBR/02 TPP LAP 17.81s · 56.7 km/h · 412 W AVG · PURSUIT/A FLY 200 9.872s · 72.9 km/h · 1456 W PEAK · NCP/03 KILO 1:00.412 · 59.6 km/h AVG · 647 W · MKW/04
The premise

Most performance data dies in a spreadsheet.

We built VeloIQ to be the layer between the sensor and the decision.

Founding principle · 2025

Elite cycling generates more data per session than any other endurance sport. Power meters, lap timing, rider position, cadence, heart-rate variability, recovery telemetry, environmental conditions. By the time a coach reconciles it on Monday morning, the moment to intervene has passed.

VeloIQ closes that gap. We capture from the meters and wearables the elite already trust, normalize the stream, and surface the laps that matter — during the recovery interval, not next week. Retrieval-augmented AI sits behind the analysis, grounded in the federation's own performance corpus rather than a generic model.

It is built first for U.S. elite track athletes, then portable to any national program that wants the same operating standard for less than the cost of a single set of disc wheels.

The system

Capture. Reason. Decide.

Three stages, one continuous loop. From the moment a meter goes live on the track to the post-session decision a coach hands back to the athlete — VeloIQ is the connective tissue.

01 — CAPTURE

Telemetry, unfiltered.

Direct ingest from the meters and wearables the elite already trust. Lap timing from track-side instruments. Power, cadence, position from the bike. Physiological signal from the athlete.

SRM · Power2Max · Quarq · Stages
WHOOP · Oura · Polar H10 · Garmin
Track-side lap timing · transponder feed
Video sync · positional · environmental
02 — REASON

AI grounded in your program.

Retrieval-augmented reasoning over your own data — not a generic model. Filter by time-segment and discipline-specific threshold. Surface the laps that matter. Compare against the athlete's own baseline, not a leaderboard.

RAG corpus per federation · isolated
Discipline-aware filters · sprint / pursuit / mass
Time-segment selection · last lap, last 10 min
Threshold detection · entry speed, V-loss, drop
03 — DECIDE

In the recovery interval.

The athlete sees what to fix before the next effort. The coach gets the same view, with the assessment and the recommended next set. The performance director gets the program-wide pattern.

Athlete view · clear, prescriptive
Coach view · same data, deeper analysis
Director view · roster trends, readiness
Audit log · every recommendation tracked
Disciplines

Built for the way your athletes race.

A sprinter doesn't review like a pursuiter. VeloIQ ships with discipline-specific filters and metric models, configurable to your program's training language.

The flying 200.

Entry speed governs qualification. VeloIQ isolates the 200m segment, computes entry velocity against the athlete's wind-up profile, and flags drift before the next qualifying effort.

Entry speed69.2 km/h · target 68.5–70.0
200m time10.412s · −0.08 vs PB
Peak power1,456 W · pos. 3 of 5 efforts
Cadence drop−4 rpm last 75m · flagged
RecommendHold one effort · gear up next set
EFFORTSEGMENTRESULT
01200m flying · gear 91.010.681s
02200m flying · gear 91.010.594s
03200m flying · gear 91.010.412s
04200m flying · cadence drop10.488s
05200m flying · standing recovery
ENTRY SPEED PROFILE · 5-EFFORT WINDOW · ATHLETE NJP/01 "Entry speed held through effort 03. Cadence drop on 04 suggests neural fatigue — consider closing the set or stepping to 92.0 for a controlled final effort."

The pursuit is consistency.

Team pursuit is won by riders who hold a lap-on-lap variance of less than two-tenths. VeloIQ surfaces the start lap and the four laps that followed, with rider-by-rider turn analysis.

Start lap21.08s · +0.12 vs target
Lap varianceσ 0.18s · target σ<0.20
Turn 1 (rider A)17.81s · clean change
Turn 2 (rider B)17.94s · 0.13 long
RecommendRider B · shorter turn next rep
LAPRIDER · TURNTIME
01Start · A → B21.08s
02Rider A — turn17.81s
03Rider B — turn (long)17.94s
04Rider C — turn17.83s
05Rider D — final 1k17.77s
4KM TEAM PURSUIT · TURN ANALYSIS · SQUAD A "Rider B held the lead 0.13s past optimal exchange point — the schedule absorbed it, but at qualifying pace this is the difference between a top heat and a re-row."

Position is the metric.

Madison, points race, scratch. The athlete's race is shaped by their position on the bunch and the lap they choose to commit. VeloIQ pairs power data with positional inference to expose the cost of every move.

Time in front 862% · target 70%
Surges >800W11 · NP 318W
Sprint laps won2 of 4 · 8 pts
Lap-out cost+47W on recovery laps
RecommendCommit lap 28 · skip lap 36 sprint
LAPEVENTNP / POS
10Sprint · 3rd298W · P6
20Sprint · 2nd324W · P4
28Solo bridge attempt412W · P11
30Sprint · 1st341W · P2
40Final sprint · 4th288W · P4
POINTS RACE · 40KM · POSITION + POWER OVERLAY "The lap-28 bridge cost 47 watts NP on the next recovery lap and dropped position to P11. The pattern repeats across three sessions — worth a tactical conversation."

The first 250 wins it.

Keirin and kilo are won in the standing start and the wind-up. VeloIQ pairs the start telemetry with the kilo split model and shows where the time is left on the boards.

0–125m7.42s · +0.04 vs split target
125–250m10.18s · on split
Peak power2,184 W · 1.4s mark
Time-to-peak1.4s · 0.1s late
RecommendGate work · neural primer next AM
MARKSPLITPOWER
125m0–125m segment7.42s
250m125–250m segment10.18s
500m250–500m segment13.71s
750m500–750m segment14.62s
1000mfinal 250m · v-drop 6%14.49s
KILO · 5-SPLIT BREAKDOWN · GATE TO LINE "Time-to-peak is trending 0.1s late across the last three start sessions. Neural primer the morning of, gate work twice a week — the first 125m is where the 1:00 barrier lives."
FOR THE ATHLETE

Your data. Your edge.

Every session you ride leaves a trail. VeloIQ makes that trail readable — to you first, your coach second, the program third. You own it.

  • A1Open your phone in the recovery interval. See the lap that mattered, the one that didn't, and the next set the coach has cued.
  • A2Watch your own trajectory across the build. Track exactly which weeks pushed peak power, which pushed consistency, which were noise.
  • A3Carry your data with you. A move between programs doesn't reset your file.
  • A4Privacy controls that match an athlete's reality — granular consent on who sees what, including the AI itself.
FOR THE FEDERATION

Your program. Your sovereignty.

VeloIQ is built to satisfy a national procurement committee on the first read. Data residency, audit, governance — by design, not as a footnote.

  • F1Federation-isolated RAG corpus. Your athletes' data is never co-mingled, never used to train a base model, never visible to another program.
  • F2Standardized vendor-security posture. Pre-built responses to the SIG questionnaire your security team already runs.
  • F3Performance director console — roster readiness, training-load distribution, return-to-train pipeline, all at the program level.
  • F4Pricing scaled to a national federation, not a software vendor's revenue model. A line in the budget, not a fight with finance.
Governance & trust

Procurement-ready, by design.

VeloIQ was architected by an IT governance practitioner before it was a product. Security review, data minimization, audit logging, athlete consent — these aren't features bolted on for the demo, they're the substrate.

01 / DATA

Federation-isolated tenancy

Your program's corpus is logically isolated. AI retrieval is grounded only in your data. No cross-tenant training. No silent migration.

02 / CONSENT

Athlete-granular consent

Each athlete owns a consent record. Coach access, federation visibility, and AI inference are separate toggles, logged and revocable.

03 / POSTURE

SIG & NIST-aligned

Pre-built SIG (Core and Lite) and CAIQ responses, SOC 2 trajectory, NIST CSF & AI RMF alignment. We arrive with the answers your security team will ask for.

04 / AUDIT

Every inference logged

Every AI recommendation is captured with its retrieved context. The federation can review, audit, and revoke any decision the system surfaced.

Built by

A team that has worked both sides of the line.

Sport authority

Jamie Staff MBE

Olympic gold medallist and former world record holder in the team sprint. Two decades of coaching across British Cycling, USA Cycling, and the international elite track community. The voice in the room every federation knows.

Platform & governance

Dr. Peter Maharaj

Twenty-five years building secure technology programs across higher education and elite sport. Olympic Technology Lead, Trinidad & Tobago cycling — Beijing 2008. Cycling Team Manager — London 2012. Technology advisor, USOPC.

Pricing

Built for a federation budget.

VeloIQ is delivered through custom federation contracts, scoped to the size and shape of your program. Two models, one principle — a clear line in the budget, not a fight with finance.

Model 01 · Scaling programs

Per-athlete seat pricing

Pay for the athletes you bring onto the platform, with volume rates as the program grows. Ideal for federations rolling out squad by squad, or building toward a full national base.

Model 02 · National programs

Flat federation license

A single program-wide license covering your full athlete base — predictable, uncapped, and simple to budget. Ideal for national programs standardizing on VeloIQ across every discipline.

Every engagement is a custom federation contract. Pricing is shaped to your program in a private briefing — talk to us about the right fit.
Engage

The next cycle is the one worth winning.

VeloIQ is in closed pilot through the LA28 build cycle. Federation engagements are by invitation. Athlete access is by program affiliation.