I build
AI-powered
products from
zero to production.
Senior Full-Stack Engineer building real-time dashboards, AI-augmented tools, and high-performance apps that ship at scale. React, TypeScript, Node.js, Python.
Products I've built and shipped.
From AI trading systems to intelligent kitchen apps — products and tools I've designed, built, and shipped.
TradeAI
A full-stack trading intelligence platform that uses Claude AI to analyze charts, detect setups, and generate actionable trade insights in real time. Processes 5,000+ tickers and delivers sub-2-second analysis.
TradePilot
An autonomous scanning engine that continuously monitors the market for high-probability technical setups. Uses AI agents to filter noise and surface only the highest-quality trade opportunities.
Kitchen Ledger
A full-stack kitchen management app that uses AI to parse recipes, track pantry inventory, and suggest meals based on what you have. Built with Next.js, Drizzle ORM, and Neon DB.
Simmerly
An intelligent meal planning platform that generates weekly meal plans, builds smart grocery lists, and adapts to dietary preferences using LLM-driven recommendation pipelines.
How I build systems.
Architecture diagrams showing how I design scalable, AI-powered systems from frontend to data layer.
End-to-end AI trading analysis: from user query to actionable trade insights in under 2 seconds.
Try the AI analysis tool.
Enter a stock ticker and see an AI-powered technical analysis — the same pipeline powering TradeAI.
Notes from things I've shipped.
Deep dives into engineering problems, architecture decisions, and lessons from production software.
How I Built an AI Trading Agent with Claude
A deep dive into architecting a real-time trading analysis system using Claude AI, including prompt engineering for financial data, streaming responses, and building a production-ready pipeline.
→Scanning 5,000 Stocks in Under 3 Seconds
How I designed a distributed scanning architecture that processes thousands of tickers simultaneously using worker queues, caching layers, and intelligent batching strategies.
→Building Production LLM Pipelines
Lessons learned shipping AI features to production: handling rate limits, streaming responses gracefully, fallback strategies, and cost optimization for high-volume LLM calls.
→Advanced Drizzle ORM Patterns for TypeScript
A practical guide to type-safe database queries, migrations, and schema design using Drizzle ORM — covering real-world patterns I use across multiple production apps.
→Next.js App Router Performance Patterns
How to get 95+ Lighthouse scores on data-heavy Next.js apps: RSC streaming, partial prerendering, route-level caching strategies, and bundle optimization techniques.
→Parsing Recipes with AI: A Technical Deep Dive
How I built a 98%-accurate recipe parsing system using structured LLM outputs, validation pipelines, and fallback heuristics for Kitchen Ledger.
What I work with.
Technologies I use day-to-day to build fast, scalable, and AI-augmented products.
How I think about systems
I focus on building systems that are observable, incrementally deliverable, and AI-augmented where it creates real leverage — not just to be trendy.
Open source & experiments.
Side projects, tools, and experiments I'm building or exploring in my spare time.
A CLI tool that uses Claude to review pull requests and suggest improvements, enforce patterns, and flag security issues automatically.
A local web interface for testing and iterating on prompts across multiple LLM providers simultaneously. Compare outputs side by side.
Backtesting engine for technical trading strategies with historical data. Run thousands of simulations and visualize performance metrics.
A minimal, fast blog starter built with Astro + Tailwind v4. Zero JS by default, perfect Lighthouse scores, Markdown content collections.
A minimal example of semantic search using pgvector, OpenAI embeddings, and a Next.js UI. Good reference for building RAG systems.
Distributed web scraping pipeline with Playwright workers, Redis job queues, deduplication, and structured data extraction using LLMs.
Let's build something worth shipping.
Open to collaborations, startup projects, and AI engineering roles. If you're building something interesting that solves real problems, I'd love to hear about it.