What you will do.
- Design local-first indexing, redaction, summarization, and retrieval pipelines.
- Evaluate model behavior around sensitive personal data and consent boundaries.
- Partner with product and security to keep AI useful, quiet, and user-controlled.
What helps you succeed.
- Deep experience with ML systems, retrieval, computer vision, or privacy-preserving AI.
- Strong instincts around evaluation, failure modes, and data minimization.
- Ability to explain technical decisions clearly to product, design, and policy partners.
How to apply.
Send a short note with the role name, your location, and two or three pieces of work that best show how you think. We care more about relevant evidence than polished career theater.
Contact recruiting