Moss
Founding ML Engineer
San Francisco, CA, US +1 · Full-time · Posted Mar 1, 2026
- Location
- San Francisco, CA, US
Remote (US)
About the job
Own the models and machine-learning systems that power Moss’s real-time retrieval, agent intelligence, and Action Layer, taking experiments into production under latency, memory, and compute constraints.
What you'll do
- Train, fine-tune, evaluate, optimize, and deploy embedding and reranking models, including multilingual retrieval models.
- Build evaluation pipelines and improve model efficiency for CPU, ARM, cloud, browser, edge, and device environments.
- Improve the Founding Agent and Action Layer, using production failures to strengthen datasets, evaluations, and models.
What they ask for
- Experience training or fine-tuning models and deploying them into production.
- Strong foundations in representation learning, information retrieval, and model evaluation.
- Strong Python skills and ability to write maintainable code beyond research notebooks.
- Ability to reason about quality, latency, memory, and compute tradeoffs.
Nice to have
- Experience with embedding models, rerankers, search relevance, multilingual models, or conversational-agent evaluation.
- Experience with CPU, ARM, mobile, or browser deployment; distillation, quantization, inference performance, or Rust.
Tools
Python, Rust
AI focus
Develop efficient retrieval and multilingual models, improve agent intent understanding and action selection, and build the evaluation and deployment systems that support real-time AI.
Model-written summary · the original posting is authoritative