The role, as published
About the company
Our client is a software company.
The role
Raydar is recruiting for this role on behalf of our client. Act as a senior individual contributor who defines and drives a research program on how machine learning systems keep improving after deployment. Collaborate with engineers to carry experimental results through to shipped features.
What you'll do
- Lead a new research program from the ground up, covering how models retain knowledge, adjust quickly and get better on their own.
- Guide technical direction as a senior contributor alongside a compact group of researchers and engineers.
- Turn promising ideas from the roadmap into running experiments quickly.
- Work with software and ML engineers to move findings into production, from building datasets through evaluation to integration into the product.
- Aim for meaningful advances and share results with the wider research community.
What we're looking for
- A doctorate, or comparable hands-on experience, in a quantitative or technical discipline.
- About 3 to 7 years doing applied research, whether at a startup, a research lab or a large technology firm.
- A research background in models that learn continuously, adapt with little data, improve themselves or build memory. Work in reinforcement learning or agent systems alone does not meet this bar.
- Practical experience with post-training of large models, including supervised and reinforcement-based approaches.
- A history of turning research concepts into working production systems, ideally in an early-stage environment.
- Papers accepted at leading machine learning conferences or journals.
- Strong coding skills in Python, a major deep learning framework and large-scale multi-machine training.
- Energy for a fast-paced startup where you own outcomes.
- Ability to work on-site at the office.
Bonus points
- Prior time at a research lab, large technology company or leading AI developer.
- Experience applying post-training techniques to agents that handle long, multi-step tasks.
Location and work model
- San Mateo, CA, United States
- On-site
- Full-time
Compensation, as the posting describes it
- Equity