Careers · Berkeley · Paris

Build the future of AI‑assisted science

AI is compressing the time between a scientific idea and a result. Join us to help make sure science remains rigorous and compounding in this new era of research.

See open roles

Working with Lightcone Research

  1. The work

    Help invent how science gets done in the AI era, from within the scientific community. Your work will be wide, touching agentic workflows, scientist–agent interfaces, HPC, and provenance.

  2. The setting

    Lightcone is rooted in leading scientific institutions — UC Berkeley and CNRS — and backed by philanthropic support, giving us the stability to pursue long-term goals and the freedom to execute quickly.

  3. The team

    We are a lean team of practicing scientists and engineers, based in Paris and Berkeley, directly in contact with the scientific communities we are building for.

  4. The moment

    Joining now means helping shape the project as a founding member of the technical staff. You'll have real ownership of some of the abstractions, tools, and norms that define Lightcone.

How we work

  • We build with rigor and attention to detail.
  • We ship quickly and iterate in contact with real use.
  • We work directly with the scientists we serve.
  • We build open-source code for open science.
  • We take end-to-end ownership of our work.
  • We keep roles broad — the mission is wide, and so is the work.

Open Roles

First openings coming soon

Berkeley · Paris

We’re finalizing our first role listings. Fill out our hiring interest form and we’ll let you know as soon as they go live.

Hiring interest form

FAQs

Do I need a scientific background?

A scientific background is helpful, but not strictly required. We are interested in strong candidates from software engineering, scientific research, machine learning, scientific computing, product engineering, open-source infrastructure, and related communities. The most important thing is that you are excited to build tools for real scientific workflows and that you care about making AI-assisted research rigorous, reproducible, and useful.

Do I need experience with AI agents?

Experience with agents, LLM tooling, evaluation harnesses, tool integrations, or execution environments is useful. But it is not the only way to be a strong fit. Candidates with strong systems, infrastructure, scientific software, product engineering, or research experience may also be excellent fits, especially if they are excited to work with agentic systems.

Are you more like an academic lab or a startup?

Neither exactly. We are rooted in scientific institutions and care deeply about research quality, open science, and long-term public-good infrastructure. But we are not organized like a typical academic project — we ship fast, build production-grade tools, work directly with users, and improve our systems in contact with real scientific workflows.

Should I apply if I do not meet every bullet?

Yes. Please do not self-reject just because your background does not map perfectly onto the role description. We are looking for people who can contribute to a broad and evolving technical mission. Strong candidates may come from several different paths, and we would rather talk with you than have you rule yourself out too early.

Still have questions? Write to us at info@lightconeresearch.org.