Join the Ensemble

Design and build next-generation AI evaluation frameworks and benchmarks. You'll develop metrics that reveal true model capabilities beyond surface-level performance, working at the intersection of measurement science and machine learning.

Requirements: PhD or equivalent experience in ML/AI; publications in top venues (NeurIPS, ICML, ICLR, ACL); strong Python and experiment design skills; deep interest in evaluation methodology.

Build autonomous agents that learn from practice and structured self-improvement. You'll bridge cutting-edge research with production-quality engineering, creating systems that iterate, evaluate, and refine their own capabilities.

Requirements: Strong software engineering fundamentals; experience building or evaluating LLM-based agents; proficiency in Python and modern ML frameworks; comfort with ambiguity and rapid prototyping.

Advance multimodal understanding and visual reasoning research. You'll work on problems at the frontier of how AI systems perceive, interpret, and reason about the visual world, bridging computer vision with language understanding.

Requirements: PhD in computer vision, multimodal ML, or related field; publications at top venues (CVPR, ICCV, NeurIPS); experience with vision-language models; strong experimental methodology.

Build the scalable evaluation infrastructure and open-source tooling that powers our research. You'll design systems that handle large-scale model evaluation, data pipelines, and developer-facing APIs that the community depends on.

Requirements: Strong systems programming (Go, Rust, or Python); experience with distributed systems and cloud infrastructure; track record of building reliable, well-documented open-source tools.

A 12–16 week research program for PhD students working on evaluation, agents, or vision. You'll collaborate closely with our team on a focused research project with the goal of producing publishable work and open-source contributions.

Requirements: Currently enrolled in a PhD program in ML, AI, or related field; strong research track record; familiarity with at least one of our focus areas (evaluation, agents, vision).