Senior Software Engineer - AI Oversight

About OpenMined

OpenMined is a 501(c)(3) non-profit foundation that brings together leading minds in AI, security, and privacy technology to develop open-source solutions that transform how data can be accessed and shared. Founded in 2017, we've grown into a global community of over 18,000 members working to ensure that privacy protection and AI innovation reinforce each other rather than undermine one another. We are driven by a simple challenge: the most valuable data for solving humanity's greatest challenges remains locked away due to privacy concerns, and we can change that through technology rather than compromise.

About the Role

OpenMined is a nonprofit building open-source infrastructure for federated AI. Our AI Oversight team is the only team in the world working on external access to frontier AI labs. We build the privacy-preserving infrastructure that lets independent evaluators run evaluations on frontier AI models without exposing the models or anyone's private data.

As a Senior Software Engineer, you'll work most closely with our Technical Lead Manager for AI Oversight, and you'll build core parts of this system and the products on top of it. You'll also work directly with organizations to deploy our software in their environments, so their data can be used while it stays private and under their control. Our software is moving from its first production deployment to several, with frontier labs and external evaluators, and we're adding senior engineers who can own big pieces of it.

This is a hands-on role for a strong general engineer: someone who writes excellent code, makes design decisions that hold up, and explains their work clearly.

This is an independent contractor engagement, roughly a full-time commitment, with an initial 12-month term we expect to renew. You set your own schedule, but we need your working hours to overlap substantially with US Eastern business hours (your local time between one hour behind and two hours ahead of ET).

Responsibilities

• Design, build and ship core infrastructure for privacy-preserving AI evaluation, mostly in Python
• Build and strengthen our secure enclave (TEE) solutions
• Extend PySyft, our open-source library, including workers and verified receipts
• Deploy and support our software in partner environments (frontier labs, evaluators, governments), and make those deployments reliable
• Help take the product from one production deployment to several, improving reliability, scale and user experience
• Write clean, tested, readable code, and review your teammates' work
• Turn ambiguous requirements into clear designs, and explain trade-offs to engineers and non-engineers
• Use AI coding tools to move fast while keeping quality high