About
I'm an engineering student who learns by building first, then pulling the system apart until I understand why it works.
I work mostly around AI systems, backend infrastructure, retrieval, and reliability, especially the quiet failures that make software look healthy when it isn't.
I enjoy turning fuzzy ideas into working products, measuring what breaks, and refining the details until the result feels simple.
01 / Experience
Backend Engineer
Oct 2025 - Jan 2026
Remote
02 / Projects
A few things I've been building.
LLM Reliability
A reliability runtime for LLM tool-calling agents, focused on catching silent execution failures before they disappear into successful-looking runs.
Reduced silent tool-call failures from 6 -> 0 across an initial 22-case benchmark.
Python · Agents · Tool Calling · Reliability · Benchmarks
Live
Retrieval Evaluation
Adaptive RAG evaluator - retrieval failure analysis, BM25 + cross-encoder reranking, and context compression.
Python · RAG · BM25 · Reranking · Evaluation
Memory Infrastructure
Chunking strategy explorer - how LLMs process and retrieve context, without abstraction layers.
Python · Chunking · Retrieval · Embeddings · Context
03 / Stack
tools that have earned a place in the toolbox
04 / GitHub
A little proof that I'm usually building something.
05 / Writing