
Prediction Market Terminal
Real-time arbitrage detection across Polymarket and Kalshi
Next.js 16 / React 19 / TypeScript / Tailwind CSS v4 / shadcn/ui
Read the case studySenior full-stack engineer, 5 years in. Python, FastAPI, and Next.js on systems where correctness is the requirement: a cross-exchange arbitrage terminal, a land-valuation pipeline, a JSON validator on PyPI that runs 116× faster than the library it replaced.
Specification

Real-time arbitrage detection across Polymarket and Kalshi
Next.js 16 / React 19 / TypeScript / Tailwind CSS v4 / shadcn/ui
Read the case studyZero-dependency position-size calculator for prop-firm traders
JavaScript / Chrome MV3 / Service Worker / WebSocket / Binance API
Read the case study
Premium token launch platform on BNB Chain
Next.js 16 / TypeScript / Tailwind CSS v4 / Wagmi / Viem
Read the case studyMost engagements land in one of these three. The overlap is the point: an AI feature nobody can operate isn't shipped, and a fast API nobody can use isn't a product.
RAG pipelines, multi-agent automations, and evaluation workflows built on LangChain and the OpenAI API. Contributor at Outlier and Turing on model training and benchmarking projects.
Scalable APIs and async ETL with FastAPI, Django, Pandas, and PostgreSQL. Processed millions of records and cut production API latency by up to 40%.
End-to-end web products with Next.js, React, TypeScript, and Tailwind on the front, Node.js or Python services behind. Clean contracts, tight feedback loops.
Server components by default, client only where state lives
Typed at the boundary, so the client cannot ask a wrong question
Async handlers, queue-backed work, no request left holding a lock
Migrations in version control, indices chosen from real query plans
Independent / Contract
Solo and lead engineer on production platforms across Web3, fintech, and real estate. Owns architecture, delivery, and operations end to end.
Client engagements
Delivery work for clients across commerce, fintech, professional services, and education, on stacks chosen by the product rather than by preference.
Freelance / Upwork
Backend and data engineering for automation pipelines, crawlers, and API integrations delivered to clients across time zones.
Deep in Python and TypeScript, comfortable elsewhere. I pick boring infrastructure so the interesting part can be the product.
A full build, an AI integration, or a second opinion on architecture. Most projects start with a thirty-minute call, and I reply to every serious enquiry within a day.