Alice Retrieval-Augmented Generation (RAG) is a proof of concept application designed to answer queries about Alice’s Adventures in Wonderland, Lewis Carroll’s timeless classic. This innovative solution leverages the power of modern AI to combine the strengths of retrieval-based and generation-based approaches. By integrating a language model with a customized knowledge store, Alice RAG can accurately and efficiently retrieve information from the book and generate responses related to Alice’s Adventures in Wonderland.
This project showcases my ability to implement advanced AI techniques, containerization, CI/CD, documentation, infrastructre as code (IoC), and web-based interaction using Streamlit, all while adhering to best practices in software development and deployment.
To run the image, you have to map port 8501
to a local port and set the GEMINI_API_TOKEN
environment variable. You can obtain a Google Gemini token (Google AI for Developers), and there is a free tier that you can use to try the API.
docker run -p 8501:8501 -e GEMINI_API_TOKEN=$GEMINI_API_KEY pmiron/alice-rag-llm:latest
will launch a Streamlit application locally, i.e. at http://0.0.0.0:8501. Below a screenshot of an interaction with the application.