I’m Pharrel Adisi, a software engineer and applied-AI builder focused on backend systems, product engineering, and tools that turn complex workflows into dependable user experiences.
Projects that reflect how I like to engineer: start with a real user problem, design the failure modes, and make the result measurable and maintainable.
An AI-assisted course-planning experience for Berkeley students. It combines structured model outputs with source-aware recommendations, automated evaluations, and deterministic fallbacks so the product remains useful when the model is uncertain.
A durable backend job queue built around PostgreSQL coordination. The system emphasizes safe claiming, retries, observability, containerized development, and automated testing—practical infrastructure for workloads that cannot quietly disappear.
A data-driven football analytics platform that turns match and player information into clear comparisons and decision-ready insights through an accessible web interface.
Supported data and development workflows for a mission-driven international organization, working across technical implementation and stakeholder needs.
Gained early experience with technology operations inside a major financial institution and learned how dependable systems support real business teams.
Python · TypeScript · JavaScript · SQL · HTML/CSS
React · Next.js · Node.js · Fastify · Zod · PostgreSQL · REST APIs
Vitest · Playwright · Docker · GitHub Actions · Vercel · Cloudflare Workers · Git
I’m a Computer Science student at UC Berkeley who enjoys moving between product questions and technical depth. I’m especially interested in teams using LLMs and modern infrastructure to improve how people learn, decide, and get work done.
I bring the perspective of an international student, experience across nonprofit and financial environments, and the habit of learning quickly by building.
If you’re building ambitious software and value engineers who care about reliability, users, and measurable impact, I’d love to talk.