Bharath Mohan
ML Engineer
Currently — in Amsterdam
Today
I build AI-powered data systems that run reliably in production. At Google (via Motivity Labs), I architect document-processing pipelines on Cloud Composer and Gemini that handle thousands of executions a day, turning unstructured PDFs into structured datasets — and shipped a Text-to-SQL agent that lets non-technical teams query databases in plain English.
Before that, five years at Thermo Fisher Scientific in Carlsbad, CA — growing from intern to Senior Data Scientist. I shipped a vector-based semantic-search API consumed by downstream teams, time-series forecasts powering operational decisions, and recommender + propensity models on AWS and Databricks.
Right now I'm most interested in building reliable, observable AI infrastructure — taking models from prototype to production without breaking trust. Based in the Netherlands · EU work authorization · open to AI / ML engineer roles.
Selected Work
- Arcane A terminal-native AI chat client written in Go — beautiful TUI, multi-provider, streams responses with first-class keyboard ergonomics.
- StyleDiff Reimagine your wardrobe with AI. Upload a look, describe the change, and Flux.2[pro] generates the new vision in seconds.
- Ark A modern AI chat app supporting many hosted providers and local Ollama models — clean UI, Python + Reflex, used daily.
- GitDone Track habits like git commits — a contribution-grid for daily habits with real-time sync, custom colors, and Convex backend.
Also Built
- Wander Weave Travel photos → AI-generated stories.
- Cerebro An AI quiz generator for any topic.
- Byte-Bites An AI recipe generator tailored to your taste.
- Product Hunt CLI Today's Product Hunt launches in your terminal.
- Chat-UI A clean chat interface for local + hosted models.
- Folio A markdown-based note-taking app, minimal by design.
Stack
I write mostly Python, SQL, and PySpark for production work. On cloud I use GCP — Cloud Composer / Airflow, BigQuery, DataProc, Document AI, Cloud Storage — and AWS (Lambda, EC2, S3). For DevOps: Jenkins, Docker, Git, and workflow orchestration. ML & data work spans recommenders, time-series, semantic search, and vector databases.
Day to day I use Neovim, Ghostty, Droid CLI, Obsidian, and a MacBook Pro. Pictures with a Sony A7R IV.
Education
-
M.S. Information Technology & Management — The University of Texas at Dallas, Richardson, TX
-
B.E. Computer Science & Engineering — Thiagarajar College of Engineering, Madurai, India