CompSci student building AI products.
I like turning ideas into working products using coding agents and iterative development. AI lets me spend less time writing boilerplate and more time understanding problems, refining ideas, and improving the final product.
Expected graduation: 2028 · Seeking internships — AI-native engineering & APM
A configurable backend for production-ready RAG systems.
FastAPI · pgvector · Supabase · Redis
Helping people explore scripture conversationally.
Next.js · Vercel Edge · pgvector · OpenAI API
A tutors-focused analytics tool transforming raw session notes into actionable AI progress insights for educators.
React · Next.js · TypeScript · OpenAI API
A privacy-first extension built to detect AI-generated content locally.
Chrome Extension · Transformers.js · Web Workers · Vanilla JS
A privacy-focused, self-hosted shared journal for couples.
Node.js · SQLite · Docker · E2E Encryption
A curated collection of practical agent skills focused on reusable workflows.
Claude API · YAML · Bash · Prompt Design
An end-to-end chronicle of building an AI tutor, spiraling into feature-creep chaos, purging 3,400+ lines of CRUD boilerplate, and designing a high-precision, headless Knowledge-as-a-Service engine.
Why I built a biblical chatbot that refuses to soften answers, the engineering behind a stateless RAG engine on the Edge, and how original-language datasets changed the game.
Consolidating a scattered 700MB library of technical docs and screenshots into a clean, searchable knowledge hub using rclone, Tesseract OCR, and an AI coding assistant.
A practical architecture for self-hosted retrieval-augmented generation using PostgreSQL and pgvector, without high cloud bills.
A structured, model-agnostic workflow for shipping production-grade software with AI agents without losing ownership of the output.
A technical walkthrough of winter break projects built under real-world serverless constraints on a zero-dollar budget.
I spent 18 months as a VP-Education in a Toastmasters club. It taught me something engineering rarely does: that ideas only matter if people understand them. Leading weekly meetings, mentoring speakers and planning sessions taught me how to communicate across different personalities. That's become surprisingly useful while building software.
