About the role
Product For GPU Managers License our runtime on your fleet For Engineering Teams Replace the LLM behind your coding agent For Product Teams Ship ROI-positive agentic products For Edge Devices Capable agents on-device Research Pricing Docs Blog Sign In Try the API for Free PRODUCT For GPU Managers For Engineering Teams For Product Teams For Edge Devices Research Pricing Docs Blog Sign In Try the API for Free Join our team Founding AI Scientist Engineering - Boston, MA / Remote Apply now
About
the role You will lead foundational research that powers our core systems and push the boundaries of what's possible with agent systems. This includes advancing our post-training system (synthetic data generation, fine-tuning, reinforcement learning), our runtime, and our evaluation frameworks. We are building agent infrastructure in a new direction, and this role is central to making that vision real. You will have ownership over your work, and you will be able to shape the direction of the company. We'd love to have you join us. In this role, you will: li>ul]:mt-1 [&>li>ul]:mb-2 [&>li>ul]:ml-6 mb-1 text-sm"> Design and advance our post-training system, including synthetic data generation, fine tuning, and reinforcement learning. Push our runtime (TIMRUN) forward to support increasingly longer running, more scalable agents. Publish impactful work and set new standards for structured decoding, inference optimization, and agent research. Develop new methods for evaluation and benchmarking of agent performance. Collaborate closely with Hongyin . You might thrive in this role if you: li>ul]:mt-1 [&>li>ul]:mb-2 [&>li>ul]:ml-6 mb-1 text-sm"> Want significant ownership and control over your work. Have orchestrated multi-node training runs. Have hands-on experience with reinforcement learning, fine tuning, or inference optimization. Have published work at major conferences in inference optimization, AI, or NLP. Are motivated to push beyond existing research paradigms and build something fundamentally new. Location We are based in Boston, MA, and are in the office 5 days a week.