Home

Welcome to Lucida!

Lucida is an open source AI platform to be used by other software developers to create Personal Assistants to meet their businesses/personal needs. We provide the infrastructure to combine various microservices (such as Image Matching, Question Answering, Digit Recognition) that are needed for your Personal Assistant. All in all, we expect our open source users to be familiar with software development in order to use Lucida as it is intended to be used.

To deploy Lucida as a web application follow these steps (https://github.com/claritylab/lucida/blob/master/tools/deploy/instructions.ipynb). This is for setting up the eco-system for Lucida. To understand how to demo this web app refer to (https://github.com/claritylab/lucida/blob/master/tutorial.pdf).

Lucida has a variety of back-end services (also called microservices) as described here (https://github.com/claritylab/lucida#overview). The app receives an input from the user through the web front-end. That input is sent to the command center (CMD) and classified to determine which set of service will be needed to serve the user input. Then the CMD sends requests to that set of back-end services using Thrift. Each back-end service and the CMD are contained within their own Docker containers in order to keep each module independent. This allows the user to create their own personal assistant by picking and choosing the back-end services they need and orchestrating them using Kubernetes.