Applied Research - Evals & Data

Prime Intellect seeks an applied researcher for a customer-facing role connecting AI evaluation, applied data, reinforcement-learning post-training, and agent systems. The role uses signals from real deployments to improve model reasoning, alignment, reliability, and performance, while translating customer needs into research and product priorities.

Responsibilities

• Work with customers to understand workflows, data sources, and bottlenecks. Prototype agents, data pipelines, evaluation harnesses, and verifiers for real use cases; validate solutions with customers and hand hardened systems to core teams.
• Design evaluations of reasoning, robustness, and agentic behavior. Capture, process, and version feedback, model traces, and reward signals; use evaluation data to identify regressions, emerging capabilities, and alignment opportunities.
• Develop reinforcement-learning and post-training approaches, including RLHF, RLVR, and GRPO, to align large models with domain-specific tasks. Collaborate with RL and evaluation teams on model alignment and reward shaping.
• Prototype workflow-automation, multi-agent, and memory-augmented systems. Extend agent frameworks and build scalable, cost-efficient distributed training and inference pipelines, with production observability and monitoring.

Requirements

• Strong machine learning engineering background and experience in post-training, reinforcement learning, or large-scale model alignment.
• Experience with applied data workflows and evaluation frameworks for large models or agents, such as SWE-Bench, HELM, EvalFlow, or internal evaluation pipelines.
• Deep expertise in distributed training or inference frameworks such as vLLM, SGLang, Ray, or Accelerate; experience deploying containerized systems at scale using Docker, Kubernetes, and Terraform.
• A record of ML or RL research contributions through publications, open-source work, or benchmarks.

The listing specifies New York, NY, USA; the benefits section separately describes flexible work as remote or San Francisco. Compensation is $150,000–$300,000 in cash plus equity incentives. Visa sponsorship, relocation support, a professional development budget, team off-sites, and conference attendance are offered.