Litmus

Evals for humans

Founding MTS

$120K - $180K•San Francisco, CA, US
Job type
Full-time
Role
Engineering, Backend
Experience
Any (new grads ok)
Visa
US citizen/visa only
Connect directly with founders of the best YC-funded startups.
Apply to role ›

About the role

We're building a founding engineering team with complementary strengths across infrastructure, product engineering, and AI research. Depending on your expertise, your work could involve:

Platform & infrastructure

  • Architect, build, and scale reliable infrastructure that powers our hiring and evaluation platform
  • Own technical decisions across distributed systems and platform infrastructure, supporting high-throughput workloads

Product engineering

  • Own product features end-to-end, from technical design through deployment and iteration
  • Work closely with engineering and talent leaders at frontier technical teams and turn recurring customer needs into reusable platform capabilities

AI evaluation & research infrastructure

  • Design RL environments, evaluation rubrics, and reward signals that expose meaningful limits in neo-human systems.
  • Develop experiment orchestration and tracking systems for reproducible evaluations across humans and agents.
  • Build meta-optimization systems that use evaluation results, human and agent trajectories, and failure analysis to iteratively improve evaluation design.

What we look for

Required qualifications

  • Proficiency in Python, JavaScript/TypeScript, or Node.js.
  • A track record of building, shipping, and operating production systems OR experience in an early stage team
  • Ability to break down ambiguous problems and move quickly from ideas to execution
  • Strong independent judgment, with a willingness to challenge assumptions and take ownership of decisions
  • Ability to learn quickly, incorporate feedback, and improve based on evidence

High signals

  • Ex-founder
  • Experience operating distributed systems, cloud infrastructure (AWS/GCP), or serverless architectures
  • Experience building high-throughput data pipelines and distributed data processing systems
  • Experience with production-quality ML training, evaluation, or research infrastructure
  • Hands-on experience with reinforcement learning, LLM fine-tuning, or post-training

You’ll work directly alongside the founders, shaping product and technical direction. If you have deep expertise in one or more of these areas and are passionate about the mission, you’d be a good fit at Litmus.

Equity offered.

About Litmus

At Litmus, we’re building evals for humans and redefining how talent is evaluated. Teams like Mercor, Wispr, Composio, Neo Scholars, Astranis, and more are already using Litmus to build their technical teams.

We’re at an inflection point. Every company in the world will soon be powered by neo-humans (agent assisted people). But building evals to measure neo-human performance is unexplored and difficult. It requires studying the limits of model & human capabilities across all sectors of knowledge work, designing realistic environments that you can profile multiple people against, and answering where a human adds value over a purely agentic system.

At Litmus, we’re solving this problem. We deliver hiring insights to companies by capturing how candidates use tools, reason, and make decisions in their interviews, giving agentic graders the evidence to evaluate candidate skillset and predict job performance.

Litmus
Founded:2025
Batch:S26
Team Size:4
Status:
Active
Location:San Francisco
Founders