March · Tennis
Know when not to ship.
A leak-safe evaluation surface across 61 chronological folds made the market benchmark—and the decision not to promote—clear.
Read the governance caseData · Quant · AI leadership
I turn uncertain signals into trusted decisions.
I lead quantitative research, data platforms, computer-vision programmes and AI products from first evidence through production. The common thread is simple: make quality visible, make decisions reproducible and build systems teams can actually operate.
10+ yearsApplied data and modelling
0 → 1 → runStrategy through operation
PhD mathematicsRigour without theatre
Jordan Moore, PhD
Lead Quant · Head of Technology & AI · CTO
Evidence in motion · 2026
Show why the model deserves trust.
Scroll through four public-safe evidence objects: evaluation scale, operational governance, leakage controls and a visible computer-vision proof point.
May · WNBA
Live is not permission.
Rolling holdouts and explicit promotion states turn a live operational shadow into a governed learning system.
Inspect the operating model
June · UFC
A model that can’t see the future.
Temporal data contracts, forward-moving windows and paired-orientation checks make leakage prevention observable.
Explore the model architecture
July · Snooker vision
Make the vision layer reviewable.
Calibration, annotation, temporal updates and a stakeholder-ready overlay become one inspectable product story.
Watch the public demoSelected systems
Leadership you can inspect.
Real interfaces, evaluation loops and operating systems—not a wall of logos or an abstract list of tools.
Computer vision · Data product
From broadcast frames to reviewable table state
A snooker perception and QA system that turns video into calibrated ball state, temporal updates and stakeholder-ready evidence.
Applied AI · Editorial workflow
From hand history to editorial story
A poker workbench that ranks moments, connects narrative arcs and keeps generated copy grounded in public session facts.
Auditable generation · RAG
Reconstructing a missing stanza with evidence
A retrieval-led workflow for exploring Old Norse text reconstruction while keeping source evidence and uncertainty visible.
Operating model · Leadership
How I lead applied-data products
A practical operating model for turning promising analysis into a governed, observable system with a clear owner and decision path.
Latest visual demo
Computer vision that tells a story.
The snooker demo makes the whole chain visible: scene understanding, ball identity, table geometry, state changes and the product layer built on top.
How I lead
Build the operating system, not just the model.
My job is to make a technical team more decisive: clear interfaces, explicit quality gates and evidence that survives the move from notebook to production.
Start with the decision
Define who acts, what changes and what evidence earns promotion before choosing the model or platform.
Make quality reviewable
Golden sets, failure modes, shadow runs and clear readouts turn “looks promising” into an accountable release decision.
Design for operation
Ownership, telemetry, runbooks and stakeholder language are part of the product—not clean-up work for later.
Current portfolio
Embedded where data meets the decision.
Lead Quant Researcher
Research governance, market and wallet signals, challenger models, monitoring and the path from backtest to live operation.
Head of Technology & AI
Sports-data annotation, QA infrastructure, computer-vision challenges and model-ready datasets.
CTO
AI sports-content architecture across ingestion, retrieval, generation, evaluation and editorial review.
Build something dependable
Need a data product that can earn trust?
I work with teams at the point where technical possibility needs to become a clear operating decision.