Developing the TomTom Intelligent Vehicle Assistant: Leveraging Generative AI and Azure Cosmos DB
summary
TomTom builds an in-car conversational assistant that connects a GPT-style model via Azure OpenAI API to vehicle systems, uses Azure Cosmos DB for short-term conversation state and personalization, and implements RAG with vector search for up-to-date local knowledge. Development moves from playground to LangChain prototypes to a cloud app deployed with Bicep, then replaces LangChain for production. Live demo resolves EV charging, routes to a restaurant, and reports arrival with 97% battery, showing the system actually works.
AI-written summary · about
chapters (4)
related talks
48mTomTom brings AI-powered talking cars to life with Azure
46mBuild and maintain your company Copilot with Azure ML and GPT-4
48mGetting started with generative AI using Azure OpenAI Service
45mThe era of the AI Copilot
20mBuilding AI Applications with Azure Cosmos DB JavaScript SDK
25mMultimodal Conversational Interfaces with GPT and Vision AI
01h 00mUnleashGPT: Mastering the Implementation of Large Language Models
16mUnlocking AI Innovation: A Conversation with Eric Boyd and Asha Sharma