Applied AI Engineer, Evals
sable · San Francisco ·
- Work mode
- Onsite
- Category
- AI engineering
Designs and runs the evaluation system for Sable's AI employee, Aidan, which leads customer calls via realtime voice, vision, and browser use. Day to day you define what a 'good call' means, build LLM-judge graders, benchmarks, and scoring harnesses, score live calls, and turn findings into runtime, prompt, and model improvements.
About Sable
Sable built Aidan, the first AI employee who can lead customer calls using realtime voice, vision, and browser use. Aidan runs a live, two-way conversation inside a real product environment, clicking through the product like a human, watching the user's screen, and adapting the journey on the fly. Every conversation feeds a self-improving context graph we call the Brain, so Aidan gets smarter with each call.
The role
You translate what makes a good performance into clear, defensible signals, and use those signals to improve Aidan. Role involves carrying out evaluations of agents on a broad range of areas on the frontier including voice, vision, and browser use, and using the findings to optimize the agent, context, and models.
What you'll do
Design evaluations for voice, vision, and browser use to direct and accelerate improvements in the agent, context, and models
Build and maintain our internal benchmarks and simulations
Work on systems to score customer-specific scenarios and identify gaps in current capabilities.
Who you are
Engineer with statistics and/or applied machine learning experience
Comfortable in a production codebase
Bonus: prior experience with evaluations or with voice, realtime, or browser use/computer use agents.