ML/AI Operations

Hop maintains ML/AI systems in an ongoing manner

Here at Hop, we build and maintain machine learning infrastructure to enable existing research and product teams to be more effective in their work.

This often involves setting up compute and storage infrastructure for both raw data and features, as well as systems to track experiments and enable reproducible research. Our operations team members are motivated by efficiency, reproducibility and productivity.

Although bespoke solutions are sometimes necessary, we prefer to assemble them from well-understood (and preferably open-source) parts: Docker, Postgres, Kubernetes/Slurm, Metaflow/Airflow, Databricks, Weights and Biases, etc. 

Featured Case Study

Toyota Research Institute’s Human Interactive Driving research team faced challenges as their experiments grew in complexity and scale. Infrastructure and operational needs of the researchers outpaced their existing capabilities. Experimental datasets were approaching hyperscale levels. They needed advanced engineering and operations support.

Are you looking to streamline and scale your ML/AI systems? Contact us to learn how Hop can help.