Product data systems

From complex data to usable product decisions.

I build systems that turn complex inputs into clear workflows, decision support, automation, reports and user-facing tools. The aim is simple: make information easier to trust, use and scale.

Why product data matters

Product data is valuable when it improves a decision, workflow, or outcome.

The work is framed around practical value: less manual effort, clearer decisions, better repeatability, and outputs that users can trust.

8h → 5minManual expert workflow compressed into a repeatable product-data pipeline.
£800 → £10Cost reduction by replacing manual build steps with automated workflow logic.
20–30%Better planning through patient-specific modelling, simulation and decision support.
~90%Faster iteration by using modelling, automation and compensation before physical work.
Interactive product systems

Software-led systems first, followed by automation, manufacturing intelligence and simulation workflows.

Each system shows the problem, the product workflow, the users served, and the measurable value created.

Swipe projects Tap a card to update details below
UrCoach Studio
Video-to-pose / 3D movement review product

UrCoach Studio

Most coaching tools either show raw video or require expensive motion capture. This turns ordinary footage into structured movement data that a coach or user can inspect.

1 camera · 1 browser · 0 markers
Input

Video frames, pose landmarks, joint trajectories, movement phases

Workflow logic

Pose extraction, joint-angle logic, confidence handling, visual overlays, 3D pose reconstruction experiments

Output

Coach-readable movement review, pose metrics, and 3D posture interpretation

Primary users

Coaches, athletes, trainers, remote rehab users

Product value

Most coaching tools either show raw video or require expensive motion capture. This turns ordinary footage into structured movement data that a coach or user can inspect.

CV as product evidence

Experience across product systems, data workflows and applied AI.

A focused view of the systems I have built, the problems they solved, and the product value created across data-rich technical workflows.

2026 – Present

UrCoach

Founder / Full-Stack AI Product Builder
  • Building browser-based movement analysis from standard video.
  • Developing pose extraction, metrics, 3D reconstruction concepts and LLM-assisted feedback.
  • Use cases: gym training, tennis, cricket, rehab and remote coaching.
2024 – Apr 2026

Imperial College London

Research Associate / Product-Focused AI Engineer
  • Built scan-to-manufacture automation for personalized NIV masks.
  • Reduced manual design from hours to minutes and lowered device cost substantially.
  • Translated clinical and fabrication constraints into a repeatable geometry/STL workflow.
2022 – 2024

Polytechnique Montréal / Philomec

Digital Planning & Simulation Engineer
  • Built digital surgical planning workflows for orthopaedic implant decisions.
  • Connected imaging, 3D reconstruction, FEA, gait simulation and reporting.
  • Made computational outputs clearer for surgical planning and decision support.
2014 – 2021

HP Labs / SUTD / University of Auckland / Max Planck

R&D Engineer across Materials, Simulation and Manufacturing
  • Built workflows for self-assembling casts, robotic printing, thermal optimization and bone remodelling simulation.
  • Combined product design, materials, FEA, fabrication and validation.
Execution stack

Tools used to move from technical data to product-ready systems

PythonReactAPIsData PipelinesWorkflow AutomationComputer VisionApplied AIAnalyticsReporting SystemsDashboardsDecision SupportValidation LogicProduct PrototypingTechnical Documentation
Core competencies

What I bring to product and platform teams

  • Data-to-workflow and data-to-product automation.
  • Technical data processing, validation, reporting and user-ready output generation.
  • Applied AI for image, video, structured data, workflow intelligence and user feedback.
  • Decision-support products that convert complex analysis into clear actions for technical and non-technical users.