How KPMG Australia Leverages Azure Cosmos DB for Generative AI
summary
KPMG stores conversational histories in Azure Cosmos DB NoSQL and keeps document embeddings in Azure Cosmos DB for MongoDB vCore to power a retrieval‑oriented generation model for their chatbot Kima. They integrate embeddings with a graph‑based HNSW vector index, which they report dramatically improves search relevance and query performance across their document set, enabling fast retrieval for generation and simplifying the data pipeline for conversational context.
AI-written summary · about
related talks
17mAzure Cosmos DB in AI Workflows: From Knowledge Retrieval to Microsoft Fabric Integrations
48mGetting started with generative AI using Azure OpenAI Service
28mClosing Keynote: Azure Cosmos DB as the Backbone for AI-Enabled Enterprise Applications
36mGENAI-Powered chatbot to increase team efficiency
24mInfuse Generative AI in your apps using
48mThe power of AI and Copilot for Azure Databases
22mGenerative AI application stack and providing long term memory to LLMs
07mCustomer Interview: How H&R Block Uses Azure Cosmos DB to Power AI and Scale for Peak Tax Season