Similar Jobs
See allStaff+ Research Engineer — Recursive Self-Improvement Lead
Sequen
Python
LLM Agents
Reinforcement Learning
Staff AI Engineer, Agents and Evals
MimoLabs
US
Python
Agent Systems
Evaluation Design
Forward Deployed Engineer
Lunon
US
LLMs
Software Engineering
Cloud
AI Engineer
Elicit
US
Python
Evaluation Design
Software Engineering
Senior Software Engineer, AI Platform
Vanta
US
LLM
Agentic Workflows
Backend Services
Your mission:
- Turn Bonsai's self-improvement loop into a product non-technical people can use.
- Own how real work becomes improvements, prove they help, and keep people in control without prompts or configuration.
Problems you'll tackle:
- Spot repeated workflows and corrections, then propose the right improvement at the right moment.
- Pick relevant skills and memories from many within a finite context window.
- Prove an improvement helped, and fix or retire it when it goes stale.
What you'll build:
- The skill runtime for discovery, retrieval, loading, and composition inside the agent loop.
- The pipeline from session history to new skills and memories, plus review surfaces for accept, edit, or reject.
- Evals, sharing, versioning, and provenance for team and marketplace distribution.
YakLabs
YakLabs is building a desktop AI application that turns knowledge workers into AI power users without docs, API keys, or knowing what to ask for. It is a small, high-intensity founding team with funding, working in person in New York City and remotely for engineering.