Machine Learning Engineer II @ Mozilla
Building the backend infrastructure that powers AI capabilities across Firefox.
I'm a machine learning engineer with a full-stack background, working across model serving, backend tooling, and the frontends on top of them. I like hard problems with simple solutions.
Right now I'm on Mozilla's AI Platform team, developing backend tooling and model hosting/serving infrastructure that powers Firefox's AI features.
Before that I built enterprise agentic and RAG systems at HGS Digital — including a back-office automation platform that cut case-worker analysis from 4 hours to 1. I also led pre-sales teams building LangGraph agentic demos, modernized a monolith into a React/Node.js architecture at Equity Sales Finance with zero downtime, and shipped TensorFlow Lite models to edge devices at Austin GIS at 80% inference accuracy.
- 🔭 Working on model serving infrastructure and developer tooling for AI teams
- 🌱 Going deeper on distributed systems and inference optimization
- 📫 Reach me at noahdelangel@gmail.com
| Project | What it is | Stack |
|---|---|---|
| CHC+ Innovations™ | AI-drafted Continuing Healthcare checklists for UK councils | Python · React · AWS |
| Automated Prospecting with LLMs | Prospecting pipeline that scrapes and qualifies client profiles with LLM-driven scoring | Python · Selenium · AWS · MySQL |
| Computer Vision Digital Signage | TFLite models embedded in an Android app for ad preemption and programmatic advertising | Kotlin · TensorFlow Lite · AWS |
| Internet Plagiarism Checker | NLP-based plagiarism detection over live web results | Python · NLP · React |
More write-ups, architecture diagrams, and case studies live at noahdelangel.dev.