Role Overview
Fluent in English with excellent communication skills, expert in PyTorch or JAX, comfortable working independently in large codebases, writes clean high‑performance Python code, self‑directed, collaborative, low‑ego, pursues advanced research experience in areas such as agents, multimodal models, robotics, diffusion models, cybersecurity, 3D understanding, time‑series analysis, contributes to open‑source or industry codebases, publishes in top venues, and enjoys improving code through testing and CI pipelines.
Responsibilities
- Drive innovative research and collaborate with clients on complex research projects.
- Develop state‑of‑the‑art models across text, image, and speech modalities using novel methods.
- Run pre‑training, post‑training, and deploy models on clusters with thousands of GPUs.
- Generate and curate data for pre‑training and post‑training, ensuring model performance exceeds expectations.
- Develop tools and frameworks for data generation, training, evaluation, and deployment.
- Collaborate with cross‑functional teams to build agents and retrieval pipelines.
- Manage research projects and communicate with client research teams.
Requirements
- Fluent in English with excellent communication skills.
- Expertise in PyTorch or JAX.
- Ability to work independently in large codebases with minimal guidance.
- Writes clean, readable, high‑performance, fault‑tolerant Python code.
- Self‑directed, navigates uncertainty without roadmaps.
- Collaborative, low‑ego, eager to learn.
- Proven track record via personal, professional, or academic projects.
- Currently pursuing a PhD or Master’s in Mathematics, Physics, Machine Learning or related field (exceptional candidates from other backgrounds welcome).
- Availability for at least 6 months, preferably nearing study completion.
- Research experience in agents, multimodal models, world models, robotics, diffusion models, cybersecurity, 3D understanding/generation, or time‑series analysis.
- Contributions to large open‑source or industry codebases.
- Publications in top academic journals or conferences.
- Enjoys improving existing code through typing, testing, and CI pipelines.