Interview Preparation Using Redis & OpenAI

Redis plays a crucial role in the LLM & GenAI wave with it's ability to store, retrieve, and search with vector spaces in a low-latency, high-availability setting. With its heritage in enterprise caching, Redis has both the developer community and enterprise-readiness required to deploy quality AI-enabled applications in this demanding marketplace.

OpenAI is shaping the future of next-gen apps through it's release of powerful natural language and computer vision models that are used in a variety of downstream tasks.

This example Streamlit app gives you the tools to get up and running with Redis as a vector database and OpenAI as a LLM provider for embedding creation and text generation. The combination of the two is where the magic lies.

Run the App

Create your env file:

$ echo "OPENAI_API_KEY=<YOUR_OPENAI_KEY_GOES_HERE>
OPENAI_COMPLETIONS_ENGINE=text-davinci-003
OPENAI_EMBEDDINGS_ENGINE=text-embedding-ada-002
REDIS_HOST=redis
REDIS_PORT=6379
TOKENIZERS_PARALLELISM=false" > .env

Fill out values, most importantly, your OPENAI_API_KEY.

Run with docker compose:

$ docker-compose up

Add -d option to daemonize the processes to the background if you wish.

Download and add the following .bin file to your repository, under the /qna directory (too large to share on GitHub): https://drive.google.com/file/d/1wKEWNrY8Wicq3Jb59uiUwOEm5oFON0wJ/view?usp=sharing

Navigate to:

http://localhost:8080/

The first time you run the app -- all documents will be downloaded, processed, and stored in Redis. This will take a few minutes to spin up initially. From that point forward, the app should be quicker to load.