Introducing LIT

Take advantage of our years of experience successfully preparing data, designing & training neural networks, and deploying trained models within your own organization. LIT is an open-source freemium platform that can be used both the cloud and on-premise that enables your team to:

  1. Reduce the cost of labeling, preparing, & staging data pipelines into high-performance streams
  2. Provide a unified web platform that allows data scientists to analyze, share, and compare results of running neural network experiments spanning multiple logical compute devices without regard to physical
  3. Containerize trained models, allowing them to be stored, version controlled, and deployed to compute devices consistent with other existing IT systems.

Our valued clients in the healthcare industry use the LIT as a cost-efficient platform for labeling their incoming data streams:

… as a shared authoring platform for their internal and external (us) data science teams to collaborate.

… as a way to easily track multiple neural network architecture experiments that running across multiple machines on a single web-accessible pane of glass.

… as a platform for continuous integration, providing an interface to audit already deployed models and feed new training data back into the learning systems.

Drop us a line to see a demo today: