Software engineering student at 42 Nice — System & AI.
Top of my cohort at 42, on the common core curriculum. I build systems that are rigorous: real-time servers, concurrent programs, and AI tooling I can measure.
Open to a software / AI internship or apprenticeship.
Focus areas: fullstack & systems programming · applied AI (LLMs, RAG, retrieval, evaluation) · concurrency
| Project | What it is |
|---|---|
| transcendence | Multi-tenant document vault with a built-in RAG assistant — 8 containerised FastAPI services behind an Nginx + ModSecurity gateway, PostgreSQL/pgvector, React frontend. |
| agent_smith | Autonomous code agent built entirely from scratch (agent frameworks prohibited): a Thought → Code → Observation loop where the LLM writes and executes real Python to solve programming tasks, using code-based tool calling instead of JSON. |
| rag_against_the_machine | Hybrid retrieval-augmented generation: dense embeddings + BM25 over ChromaDB, with measured retrieval metrics (Recall@k) and documented trade-offs. |
| the_answer_protocol | Real-time multiplayer server & clients in Rust — custom TCP protocol, async concurrency with Tokio, and a GUI client built with the Bevy engine (ECS). |
| codexion | Concurrent systems in C — dining philosophers with pthreads, deadlock prevention, and a custom scheduler. |