Chat with Podcast Content

This project uses advanced AI tools to provide semantic search and a chat interface for analyzing and interacting with podcast transcript files. By integrating powerful AI technologies, this tool enables users to glean valuable insights and conduct meaningful searches through podcast content.

Features

Semantic Search: Search podcast transcripts with high precision using Cohere embeddings. Enables contextual understanding, beyond simple keyword matching.

Chat Module: Chat with podcast transcripts using Cohere's CommandR Plus language model. Ask specific questions and receive insightful responses derived from the podcast content.

Technologies Used

1. Programming Language: Python
2. Embeddings: Cohere Embedding v3
3. Vector Database: ApertureDB
4. Chat LLM: Cohere CommandR Plus

Setup Instructions

Prerequisites

1. Python 3.8 or later
2. A Cohere API key (for embeddings and chat functionality)
3. ApertureDB setup (refer to ApertureDB documentation: https://docs.aperturedata.io/Setup/client/adb)

AI Chronicles podcast links:

YouTube: https://www.youtube.com/channel/UCVXhvCHU_wZ7lnzal-ZpfMQ/

Spotify: https://podcasters.spotify.com/pod/show/sonam-gupta55

This repository will contain just the Python files and not the actual dataset. For the detailed process, please check out my Substack blog: https://open.substack.com/pub/sonamcoffeenlp/p/semantic-search-to-glean-valuable-deb?r=pon7a&utm_campaign=post&utm_medium=web