Powering Vector Search with Real-Time Data

Speakers: Zain Hasan (Senior Developer Advocate at Weaviate), Merlin Carter (Developer Educator at Quix).
Host: Kris Jenkins.

RAG (Retrieval-Augmented Generation) has gained popularity for enabling language models to access data beyond their training set, such as private or recent data. However, many applications require data that can only be retrieved in real-time or benefit from the most current information available. In this webinar, we will explore two use cases where real-time data enhances vector-powered retrieval applications, including RAG and product recommendations.

1. RAG on a Live Transcription of the Session.
We will demonstrate how to ingest a live transcription into Weaviate using microphones and a speech-to-text API. Attendees will see how a chatbot can answer questions about topics just discussed. This capability is particularly useful for those who join meetings or webinars late and need to catch up quickly.

2. Real-Time CDC on a Product Inventory Database.
We will show how to update a product catalog with new entries and incorporate these changes into your vector database in real-time. Using Change Data Capture (CDC), we will listen for updates and ingest new text descriptions and images into Weaviate. We will then demonstrate how these real-time updates enhance the product search experience.

Links to the repos featured in the demo:
https://github.com/quixio/template-ra...
https://github.com/quixio/template-ve...

Live demo UI - search the conversation:
https://quix-weaviate-webinar-chat-pd...