SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Salary: USD 150,000 - 300,000 / annual
Prime Intellect is building the open superintelligence stack—infrastructure for frontier AI labs to train and deploy post-trained models at scale. The company has raised $150M from Founders Fund, Radical Ventures, NVIDIA, and prominent AI operators including Andrej Karpathy and leaders from Perplexity, OpenAI, Databricks, and others.
In this Applied Research role, you will build and post-train agents for autonomous scientific research. You'll design long-horizon environments where agents search literature, query scientific databases, run code and domain tools, and select experiments within compute or experimental budgets. Working directly with research partners, you will turn their workflows, datasets, and outcome criteria into trainable tasks, evaluations, and reward signals.
Key responsibilities include:
- Building long-horizon environments for agents to conduct autonomous research across scientific domains
- Converting research partners' workflows, datasets, and success criteria into trainable tasks and reward signals
- Designing verifiers for correctness, novelty, and reproducibility; detecting label leakage and unsupported conclusions
- Connecting programmatic checks, simulations, domain models, and expert feedback into training loops
- Developing RL and multi-agent methods for hypothesis proposal, testing, and critique; comparing against strong baselines
- Ensuring reproducibility of code, data, logs, and configurations; releasing reusable environments and evaluation frameworks
You will maintain a substantive research agenda while being externally facing—most researchers work directly with customers, users, or research partners.
REQUIREMENTS:
- Depth in reinforcement learning, agents, or post-training systems, with ability to acquire relevant scientific context
- Hands-on experience building and debugging model-training, post-training, agent, evaluation, or closely adjacent experimental loops
- Strong experimental judgment: meaningful baselines, careful trace inspection, clear failure attribution, willingness to change direction based on evidence
- Strong Python and ability to build artifacts that other researchers and engineers can run, inspect, and extend
- Comfort owning ambiguous problems and keeping partners and teammates informed as you resolve them
- Prior work in autonomous research, ML for science, scientific computing, or reproducible computational research is useful but not required
- Evidence of exceptional work valued: publications, open-source systems, benchmarks, shipped products, or difficult projects demonstrating learning from failure
