About the Company/Team

We’re a newly founded, venture-backed startup built by a team from MIT, UC Berkeley, and Harvard Business School. We’re building next-generation multimodal systems that move people from passive watching to active, interactive experiences—and we already have traction with large gaming platforms. We’re hiring an engineer to productionize and maintain our ML stack, from training to deployment to demos.

About the Role

This role focuses on supporting applied machine learning research and product development by deploying and versioning models with Hugging Face tools such as Transformers and Spaces, managing datasets through curation, storage, and versioning, and building interactive demos and LLM-powered data pipelines for tasks like generation, filtering, deduplication, and evaluation. The ideal candidate has strong Python and PyTorch skills, hands-on experience with Hugging Face model deployment, and the curiosity to use emerging tools and agents to tackle unfamiliar problems. Additional strengths may include experience with streaming or interactive generation, model distillation or acceleration, PyTorch Lightning, and prior product or internship experience.

In return, the role offers hands-on research exposure alongside top researchers and the potential to co-author a paper at a leading machine learning conference.