Design · Build · Deploy · Operate
Work
Behavior Classification Inference Pipeline — 4.4× Throughput
Reworked the production core package that extracts 903 features from 6-axis IMU raw signal, for 4.4× throughput and 3.3× end-to-end speedup.
4.4×throughput24h real-animal · 2.16M raw rows
423win/sec8,640 windows · was 96
208 / 0pass / failfull suite · output-equivalence verified
Internal LLM Data Agent Platform
An assistant that answers questions about internal data in natural language. Moved from MVP to a tool-calling agent, with the architecture decided by A/B measurement.
5 / 5compound queriesown eval set · old structure 0/5
730schema documents95 tables · 635 columns
4surfacesMCP · Slack · REST · app API
Company Knowledge Base — Privacy-Safe Compilation Pipeline
A knowledge base where an LLM compiles and maintains internal conversations. Answers come from compiled documents, not from retrieval over raw text.
2de-identification layersregex + GPU NER
12ingest channelsinternal messenger
Threat Intelligence Data Platform — Ingest to Serving
A TI platform run end to end by a team of three: crawling, storage, search, API and UI. This is where I learned the difference between building a model and running one as a service.
6stages ownedingest → clean → store → search → API → UI
3team sizeinternal analysts + external customers