Patrick Semler

Texas A&M CS ’28 · College Station, TX

patricksemler@tamu.edu

I'm a Computer Science student at Texas A&M, currently doing machine learning research on veterinary diagnostics. I build full-stack AI systems at Apply Finch, and before that wrote backtesting and risk infrastructure at Maroon Fund. On my own time I'm building LeetMind, an adaptive coding-practice platform that aims each problem at what you are actually weak at, and Phobos, a self-hosted AI chief of staff that can safely write and deploy new tools for itself.

Projects

  • LeetMind

    An adaptive coding-practice platform that generates original problems tailored to each user's per-concept mastery, so every session lands at the edge of their ability. Each generated problem clears a six-stage gate — schema, compile, differential, boundary, example, and mutation checks — before it is ever shown, and the judge holds roughly 75 submissions per minute with zero incomplete jobs.

    • TypeScript
    • Python
    • React
    • Fastify
    • PostgreSQL
    • Docker
  • Phobos

    A self-hosted AI chief of staff for one person — 63 natural-language tools across 13 capability modules, reachable from a Telegram bot and a Realtime-synced React dashboard, that tracks habits, watches an inbox and calendar, and messages first. A self-extending forge plans and generates new TypeScript modules in isolated Git worktrees, then gates them behind typecheck, tests, and an independent cross-vendor review; nothing reaches production without human approval, and anything the system writes to itself is one command from a revert.

    • TypeScript
    • Node.js
    • React
    • Supabase
    • MCP

Experience

  1. Machine Learning Research Assistant

    2026-06present

    Texas A&M University

    Diagnostic classification across 144 veterinary conditions — cut the production classifier's peak memory by 83% while holding 91% accuracy, and grounded the Gemini RAG pipeline in full case context so it stopped inventing care plans.

    • Python
    • scikit-learn
    • Gemini
    • RAG
  2. Full-Stack Developer

    2026-01present

    Apply Finch

    An AI job-application platform built with a 5-developer team — routed pipeline stages across cost-optimized LLMs to get generation under $0.01 per application, and hardened the async pipeline against 4 failure modes it used to lose work to.

    • React
    • TypeScript
    • Flask
    • PostgreSQL
    • AWS
  3. Quantitative Developer

    2026-022026-05

    Maroon Fund

    Backtesting and risk infrastructure — added Jensen's alpha, OLS beta, and tracking error against a 3-source SPY fallback, and raised backtest fidelity by modeling slippage, commissions, and forced exits.

    • Python
    • Pandas
    • Plotly
    • pytest