Software Engineer, Agent Harness

LM Studio•New York, NY

5d•$150,000 - $350,000•Hybrid

About The Position

LM Studio is used by millions of people around the world to run AI on their own computers, and now with Bionic - also in the cloud. Our values prioritize putting the human in the center, and creating tools that we want to use ourselves, and recommend to our friends and family. As a team, we work with high technical intensity and personal responsibility. We are looking for curious, self-motivated, creative, and technically excellent teammates to join us and build the future of human-AI interactions in software. The Role We are looking for an Agent Harness Software Engineer to build the core agent system behind Bionic, LM Studio’s agent for work and code. You will make open models effective at working with codebases, documents, files, software tools, and the user’s computer. Open models vary widely in how they reason, call tools, use context, and recover from mistakes. You will build the software that handles those differences and gets the best possible agentic performance from models running locally, on linked devices, and in LM Studio's inference cloud.

Requirements

Significant experience building production software involving language models, developer tools, or automation
Strong understanding of tool calling, structured model output, context construction, and streaming inference
Experience building stateful systems that perform multi-step work and recover cleanly from failures
Excellent TypeScript and general software engineering fundamentals
Deep interest in open models and a practical understanding of their strengths and limitations
Works with high technical intensity and takes personal responsibility for the finished product

Responsibilities

Build and improve the agent harness that powers Bionic
Make local and cloud models reliable at using files, search, Git, the shell, web tools, and document tools
Integrate new open models and optimize how they perform in the Bionic agent harness
Manage conversation history, working context, tool results, interruptions, and recovery
Build repeatable evaluations and diagnostics for understanding where and why an agent fails
Constrain the model in a smart way so it remains useful
Work across Bionic, the LM Studio runtime, and LM Studio cloud to improve the complete system