# Research Engineer, Interpretability

Hiring organization: [Menlo Ventures Portfolio](https://career.thegoodapps.co/organizations/menlo-ventures-portfolio)

Canonical page: https://career.thegoodapps.co/jobs/c9519593-3e61-4e9e-aa5d-803a14048068

Listed on Menlo Ventures Portfolio's own careers site. Applications go to them directly.

- Employment type: full time
- Seniority: Senior
- Location: San Francisco, CA
- Salary: 315000 – 560000 USD per year

## Summary

This role builds the computational infrastructure and tooling that lets researchers understand how large language models actually work internally, rather than treating them as black boxes. It's suited to experienced software engineers who want to work on AI safety through hands-on infrastructure work, enjoy translating research needs into systems, and are comfortable optimizing across the full stack from GPU kernels to user-facing tools.

_Our summary, not Menlo Ventures Portfolio's wording._

## Skills named

CUDA, Go, Java, JAX, Python, PyTorch, Rust

## Required

- 5-10+ years building software
- Proficiency in at least one of: Python, Rust, Go, Java
- Productive with Python
- Ability to quickly learn unfamiliar technical domains
- Strong prioritization and ability to work with ambiguity
- Collaborative working style
- Bachelor's degree or equivalent

## Nice to have

- Experience optimizing large-scale distributed systems
- Language modeling fundamentals with transformers
- High-performance LLM optimization (memory, compute efficiency, parallelism, throughput)
- Hands-on experience with PyTorch/CUDA on GPUs or JAX/XLA on TPUs
- Experience building tooling for research teams or conducting research with complex engineering challenges
- Background in interpretability or AI safety research

Apply on Menlo Ventures Portfolio's site: https://job-boards.greenhouse.io/anthropic/jobs/4980430008