Developed an AI agent to support an automated bug-fixing initiative, continuously addressing ~7% of bugs org-wide, with 60% of generated code accepted without modification. (Python, Hack, Confucius (LangChain equivalent), RAG, MCP)

Led engineers in building a scalable device setup platform that cut device bring-up time from 4 weeks to 1 day and unlocked a new business path for third-party partners, enabling self-serve device onboarding. (TypeScript, React Native, GraphQL)

Modernized an observability platform and integrated automated AI-powered Root Cause Analysis, driving organization-wide adoption through federated implementation across partner teams, closing a 30% telemetry-data loss gap and cutting metrics drop investigation time from ~2 weeks to ~2 days. (TypeScript, React, Python, SQL)

Drove a 147% increase in team AI code-generation usage by creating presentations on effective AI usage, supporting broader organizational AI adoption goals.

Conducted technology consulting, full-stack web development, and enterprise application development across diverse organizations, from startups to large corporations.

Selected clients:

● An AI research organization

Established a 25-person cross-functional team to build a system to create critical data products (text corpora) for the development of LLM interpretability. The system enabled arbitrarily many data producers to cohesively contribute to a singular body of work, delivering a 5-million-word text corpus in 6 months. To ensure data quality and diversity, built a peer-review system, a trust-based promotion system, and a tiered payout system, resulting in a mean self-BLEU score of 0.11. (Python, PyTorch, AWS, Integromat)

● An e-commerce company with millions of users.

Built a product shipping/return management system and an AI recommender system, increasing subscription renewal rates by 21%. (Python, PyTorch, Flask, Redis, Postgres, AWS)

● A location-based mobile ads startup

Led development of the company’s pre-pivot flagship product—a unique location-based physical

commerce concept, from data model design to product launch.

● A climate change international NGO

Developed a location-based interactive digital media web app for Al Gore’s climate change non-profit, which was launched during a global fundraising event to millions of users, and handled a peak load of 1000s of concurrent users. (TypeScript, React, Ruby, Redis, Postgres, AWS)

Recruited a team of 5 engineers and led the team through the end-to-end development of the company’s flagship product: a web and mobile technical talent marketplace, shipping 2 months ahead of schedule. (JavaScript, React Native, Ruby on Rails, Redis, Postgres, AWS)

Ideated, architected, and personally implemented a real-time interactive SMS chat system, a calendar booking system, an automated voice conferencing, an internal candidate search engine, and a subscription payments system. (Ruby, Twilio, GCal, Elasticsearch, Stripe)

Developed a web application for the computation of risk analytics using Monte Carlo simulations, and the programmatic generation of PowerPoint presentations. Developed a multi-platform data pipeline, where data is extracted from a data warehouse, transformed with XSLT, and loaded onto a Windows server. (Java, ASP.NET, C#, DB2, MSSQL Server)

Developed a desktop application for trading teams to size and analyze option trades. (Java, DB2)

Developed a tool to measure database API performance and utilized the tool to identify and eliminate performance bottlenecks, reducing average application query time by 33%. (DB2)

Developed a flexible data extraction app, allowing business units to self-service data requests, increasing engineering team bandwidth by ~15%. (Java, Spring, Hibernate, DB2)