ML Compression Engineer

The Compression Company•San Francisco, CA

About The Position

Build the compression technology at the core of The Compression Company. Design and train neural compression models, improve compression ratio while preserving quality and downstream utility, and make the training system fast and repeatable so the company can produce state-of-the-art codecs for new datasets with minimal manual effort. This work directly drives product performance and the pace at which the company can ship into new customer environments. The role involves improving rate/quality performance for real customer datasets, building automated training runs that explore model and training settings and select strong candidates, extending evaluation to reflect real user needs beyond offline metrics, and improving reproducibility and speed across training and benchmarking. The Compression Company builds dataset- and deployment-specific codecs as production-ready encoder/decoder artifacts that run on hardware ranging from edge devices to cloud GPUs. It is a small, highly technical team focused on ownership, shipping real systems, measurable performance, engineering discipline, and a platform that improves with each deployment.

Requirements

  • Experience training and shipping machine-learning models end-to-end.
  • Care about reliability, performance, and clean experimentation.
  • Strong PyTorch skills.
  • Comfortable owning training and evaluation.

Nice To Haves

  • Enjoy turning research work into usable systems.

Responsibilities

  • Design and train neural compression models.
  • Improve compression ratio while preserving quality and downstream utility.
  • Build and improve fast, repeatable training systems for producing codecs for new datasets with minimal manual effort.
  • Improve rate/quality performance for real customer datasets.
  • Build automated training runs that explore model and training settings and select strong candidates.
  • Extend the evaluation suite to reflect real user needs, not only offline metrics.
  • Improve reproducibility and speed across the training and benchmarking loop.

Benefits

  • Founding equity with meaningful ownership.
  • Full relocation and visa support.
© 2026 Teal Labs, Inc
Privacy PolicyTerms of Service