Project Fleming

Introducing Project Fleming: Originally devised for Shell, this model-agnostic, AI-powered Discovery Tool harnesses open source models from Hugging Face on Small CPU Clusters on Databricks. We invite contributors and users alike to join us in enabling Discovery for all organizations looking to power their Inner Source initiatives or use Artificial Intelligence for specialized Discovery tasks.

Inner Source refers to the deployment of Open Source practices of code and information sharing within a single organization. By sharing, we reduce duplication, free up cognitive space for innovation, and increase the speed of delivery.

However, simply setting repositories to “open” is not enough to ensure effective sharing. Traditional search tools often have a narrow focus and rely on specificity.

The solution is a Discovery tool, allowing developers to find unexpected yet helpful solutions and information. Leveraging the power of Artificial Intelligence, this tool is designed to be resource-efficient, using clever programming techniques to serve models with Databricks on a CPU cluster instead of a GPU one.

Join us in contributing to Project Fleming and help drive innovation and efficiency through enhanced discovery and sharing.