SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Salary: USD 255,000 - 300,000 / annual
Notion is building a collaborative AI workspace where teams and agents think together. The Model Capabilities team owns the model layer of Notion AI, responsible for integrating frontier models as they ship, keeping inference reliable and economical at scale, and building new capabilities that other teams leverage.
In this role, you will:
- Bring new frontier models into production quickly, making them available to users and engineers
- Ensure inference reliability through better error categorization, self-healing retries, and cross-provider failover
- Own observability for the model layer, driving down both time to detection and time to fix
- Build new model-level capabilities and help product teams adopt them
- Act as connective tissue across Notion's AI teams: identify gaps, unblock people, and ensure fixes land with the right owner
You'll work from the San Francisco or New York City office on Mondays, Tuesdays, and Thursdays (Anchor Days), with flexibility on other days.
QUALIFICATIONS & REQUIREMENTS:
- Comfort with a rapidly evolving frontier: The model landscape changes monthly. You stay current, form your own informed views, and remain pragmatic about what's worth adopting now.
- AI fluency as a practitioner: You use AI tools seriously in your own work and have strong opinions about where they help and where they don't.
- Production AI systems experience: You have built and operated systems on top of LLM APIs (not just prototyped). You understand token economics, streaming, tool calling, context limits, and how each fails in production.
- Measurement before conviction: You don't ship prompt or model changes on intuition. You define what "better" means, build or use evaluations to test it, and are willing to be proven wrong by your own data.
- Cost and performance instincts: You treat latency and dollars as product qualities, not afterthoughts, and can reason about tradeoffs between them and output quality.
