Tri Do

I work on the layer between a model and the people who depend on it: retrieval, serving, evaluation, and the telemetry that makes all three arguable. Every system here was built against a ceiling I could not raise. A 6GB GPU picked the quantization, the reranker, and the vector store. $290 of secondhand homelab hardware picked the cluster topology. A zero dollar API budget picked the rate limiter before it picked the provider.

A fixed ceiling is also what makes the work measurable, because it turns most design questions into arithmetic. I try to answer those honestly. On the project below, a six arm retrieval ablation produced a clean set of per stage improvements, and every one of them sat under the smallest effect that run could actually detect, so I withdrew them. The numbers on this site are the ones that survived that.

I came to Purdue for statistics and added computer science in my second year, which turned out to be the whole story. I liked the math and I liked the systems, and AI is the place where those stop being two separate interests. I moved toward it later that same year and have not looked seriously at anything else since. Infrastructure came later and from somewhere else entirely, and that story is on the homelab page where it belongs.

The domain name, pure-white-ash.com, comes from volume 14 of Ashita no Joe:

That burning sense of worth and completeness is something I've only tasted on a bloodied ring. And this burning sensation isn't a momentary sputtering that other people my age might feel. It's so bright it burns your entire body up in an instant. And when it's over, only white ashes remain. Not even any tiny cinders. Only white ashes.

Stack

Languages

Python, Java, C, SQL, Bash

ML and LLM systems

PyTorch, Hugging Face, Qdrant, QLoRA fine tuning, GGUF quantization, hybrid retrieval, LLM as judge evaluation

Backend

FastAPI, Spring Boot, REST, GraphQL, PostgreSQL, Redis, ActiveMQ, Docker, Kubernetes, AWS, Linux

Observability and CI

OpenTelemetry, Prometheus, Grafana, Jaeger, GitLab CI/CD

Muay Thai

Two years of Muay Thai and kickboxing. I came for the conditioning, stayed for the intellectual game, and keep going for reasons I have stopped examining.

Literature

I read mostly Russian and French classics. It is where the project above got its name and most of what I think about outside work.

Manchester United

Thirteen years a United fan. I arrived for Ferguson's last title and got the decade of despair that followed. Still counting.

Projects

• Mitya

  A grounded question answering system over Dostoevsky, on one 6GB GPU. Hybrid retrieval, a two provider gateway, and a CI evaluation gate that withdrew several of my own findings.

• Homelab

  Twelve self hosted services on a $290 box, reachable only through a WireGuard tunnel and mutual TLS against a private CA.

• Other

  kurono, better-ls, better-library. Smaller things, built because the existing tool annoyed me.