Print this page to PDF for a single-page resume.
Nadja Rhodes
Software {Machine Learning} Engineer
Work Experience
Spotify · Washington, DC Jul 2025 – Present
Senior Machine Learning Engineer
- Built LLM-as-judge evaluation and prompt-testing tools for Spanish-language and text-input requests in Spotify's AI DJ, directly informing the launch decision.
- Designed evaluation and safety-review tooling combining LLM judges and deterministic checks for ~1.4B AI-generated listening reports in 2025 Wrapped, reaching hundreds of millions of users. [Blog]
- Led AI-generated music context notes from concept to test readiness, owning architecture, evaluation, and safety; batching the costliest stage cut LLM costs by over a third.
- Train and evaluate audio representation models and build tools to explore new approaches to music discovery.
ASAPP · New York, NY Jan 2019 – Jul 2024
Staff Machine Learning Engineer (2021–2024); Lead Machine Learning Engineer (2019–2021)
- Led three ML engineers, with research and product partners, to launch the company's first agentic customer support bot for a Fortune 100 client.
- Directed a team of four building a shared ML training framework used by all ~20 company models, with standardized pipelines, remote compute management, dataset versioning, and a model registry.
- Led rollout of a named entity recognition service for the company's first voice client, lifting entity card hover rates 11%.
Microsoft · Redmond, WA Sep 2013 – Sep 2018
Software Engineer II (2016–2018); Software Engineer (2013–2016)
- Developed a machine learning prototype for detecting actionable tasks in text; co-invented two granted patents covering contextual privacy and synchronization between messages and digital notes.
- Led development and public launch of OneNote.com/notebooks (4M monthly active users, 58K daily visits); rebuilt and open-sourced the OneNote Web Clipper, raising its Chrome Web Store rating from 2.73 to 4.44; led GDPR export compliance.
Education & Training
Recurse Center · Brooklyn, NY Winter 2024
Self-directed programming retreat
- Built interactive visualizations, creative coding projects, and simulations with LLM agents, learning alongside peers through pairing, demos, and study groups. [Program]
OpenAI · San Francisco, CA Summer 2018
OpenAI Scholar
- Selected for a competitive scholarship to study deep learning with Dr. Natasha Jaques, focusing on language modeling; developed and open-sourced deephypebot, a model that generates music commentary on Twitter. [Code, Demo Day talk]
Stanford University · Stanford, CA 2009 – 2013
B.S. Computer Science
- Information track; interned at NTT Communication Science Labs (2012), Google (2011), and Stanford Research (2010).
Technical Expertise
AI evaluation: LLM evaluation, human review, safety checks, release criteria, model observability, and monitoring.
Applied ML: Conversational AI, language modeling, NLU (intent, entities), audio embeddings, and vector search.
ML systems: Shared training platforms, distributed training, data pipelines, model serving, deployment, and experiment tracking.
Tools: Python, Java, TypeScript, SQL · PyTorch, Hugging Face · Braintrust, DSPy · Claude Code, Codex, MCP · Ray, Kubernetes, Flyte, MLflow · GCP (BigQuery, GKE), AWS · FastAPI, React
Conferences: AfroTech (2026) · RenderATL (2023) · ACL (2019) · NSBE (2018) · Grace Hopper (2016) · YC Female Founders (2014)