AI/ML engineer, full-stack builder, ex-acoustician.

Technical Co-Founder

An early-stage startup building an AI agent team for small businesses' online presence. Two-person founding team.

Co-founded with Kukesh Kodess.

• —Building the product end-to-end with my cofounder: brand-strategy agents, a Content Studio GenAI pipeline, AI onboarding, and a website builder that produces brand-aligned sites from a single intake.
• —Multi-provider LLM orchestration across Claude, Gemini and OpenAI using the Claude Agent SDK, OpenAI Agents SDK and Vercel AI SDK, with provider failover and structured-output validation.
• —Across the full Next.js App Router + Supabase + FastAPI stack: schema, Postgres RLS, Supabase Auth, Stripe billing, Vercel deployment.

Next.js · TypeScript · Supabase · Postgres · FastAPI · Claude · Gemini · OpenAI · Vercel · Stripe

Partner & Engineer

A forward-deployed AI consultancy — a wing of Modal — that I run in partnership with Kukesh Kodess. We embed with a team, find the workflow where AI will actually move a number, and ship the agent or tool that runs it on their own stack.

• —Forward-deployed engineering: a KPI-and-constraint diagnostic to pick the highest-ROI workflow, then designing, building and deploying the agent, model or internal tool into the client's existing stack — one production workflow live before scaling.
• —Hand off with a human review loop, training and a runbook; the client owns the code and the IP.

Full-Stack AI Engineer

2-engineer product team. Product sunset May 2026.

• —Built and shipped an AI image platform for realtors and homeowners — virtual declutter, staging, renovation and AI video walkthroughs — end-to-end to production, with a realtor in early adoption.
• —Designed async generation pipelines integrating Fal.ai, Gemini and Firecrawl with job queuing, result caching and credit-based Stripe billing.
• —Shipped the Next.js + Supabase + Postgres stack front-to-back: auth, multi-tenant data, file storage, image-diffing UI and Vercel deploys.

Next.js · Supabase · Postgres · Python · Fal.ai · Gemini · Firecrawl · Stripe

Mentor & Program Lead

Launchpad's applied-AI fellowship. Returned to run and mentor cohorts after completing the fellowship myself in 2021.

• —Mentored cohorts of fellows through real-world ML/LLM projects end-to-end — problem framing, approach and code review, evals, and shipping a working system — translating production experience into hands-on teaching.
• —Ran the program side: guiding teams, unblocking technical decisions and setting the bar for what 'production-ready' means.

Data Scientist — Member Modelling & Marketing Analytics

• —Built LTV models, engagement scoring and member segmentation across Nike's ~300M-member base on PySpark and Databricks.
• —Designed an Engagement Quality Score and affinity-overlap metrics adopted by the Member Modelling team.

PySpark · Databricks · SQL · Python · AWS (S3, RDS, Lambda, Batch)

Data Science Fellow

• —Built a medical computer-vision system for fetal-gender masking in ultrasound imagery (YOLOv5, OpenCV); published on arXiv.

Principal Acoustic Consultant · Founder

A prior career: 300+ projects across the UK, NZ and Colombia; founded an acoustic-panel manufacturer; built Python tooling for acoustic modelling.

University of Bath, UK

Dissertation: On the Confidence Calibration of Foundation Language Models.

University of Salford, UK

University of Warwick, UK