Developers and businesses can leverage low latency Nebula LLM and Embeddings model and MongoDB Atlas for building RAG use cases on voice data:

• Q&A in context of live conversations

• Sentiment and Emotional analysis on audio/video transcripts

• Semantic search across transcripts and documents

Here is how you can use our models and MongoDB Atlas to build an end to end RAG pipeline:

• Nebula Embeddings: Transform large volumes of audio and video transcripts into vector representations

• MongoDB Atlas: Use MongoDB Atlas to store these vectors and Atlas Vector Search to extract contextual information based on the user query

• Nebula LLM: Feed the contextual information to Nebula LLM for generating responses

With Symbl and MongoDB, you can drive actionable insights, analytics, and trends from large volumes of customer conversations across domains such as sales and customer service.

Core competency: Database; Stream processing; Vector Search