End-to-end-rag-app 1. Took a blog post on LLM, scraped it using beautifulsoup. 2. Took the content & split it into chunks to feed it to db. 3. Used Hugginface Embeddings to convert the chunks into vectors. 4. Used cassandra(AstrDB) to store these these embeddings. 5. Used Mixtral model & our custom prompt template. 6. Finally, created a retrieval chain and invoked it to get results.