SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Handshake's mission is to organize expert human knowledge to advance the AI economy. The company works directly with frontier AI labs on their most consequential data, evaluation, and post-training challenges, building systems that turn expert knowledge into data and evaluations that improve frontier models.
As a Member of Technical Staff focused on Coding Agents, you will help define how frontier coding agents are benchmarked and improved. This is a broad, high-ownership role for researchers who build. You will partner with world-class AI researchers and thousands of software engineers to develop new benchmarks, reward and verifier systems, agent-evaluation methodologies, and data-quality techniques centered on AI applications for coding.
Key responsibilities include:
- Creating and publishing coding benchmarks that challenge state-of-the-art frontier agents
- Understanding and shaping the frontier of coding agents, publishing studies of interesting and impactful findings
- Helping the world's top AI labs improve their models for coding
- Publicly contributing to the field through benchmarks, open-source tools, research, and technical writing
You will work on fundamental questions at the center of frontier AI progress: what should be measured, how to design evaluations that reflect real capability, how to create high-signal feedback, and how to build the environments and data systems that make those answers actionable. Early team members will have unusual influence over technical direction, operating culture, and the open-source software, benchmarks, and research products the team builds.
The company values demonstrated research capability, technical judgment, and a builder's mindset over specific titles, degrees, or career paths.
REQUIREMENTS:
- Extremely passionate about building software with agents and regularly tinkering with the latest coding agents
- Comfort operating in an ambiguous, fast-moving environment with substantial ownership
- At least one of the following:
* Previous experience working on a coding benchmark
* Published research on AI for Software / Code-Generation at top-tier venues (NeurIPS, ICLR, ICML, COLM, etc.)
* Extensive software engineering experience or strong GitHub profile of open-source contributions, along with deep knowledge of agent harnesses
