Inspired by this example from Denise Schlesinger, This Jupyter notebook demonstrates the use of AzureOpenAI's GPT-4o model to generate a comprehensive natural language description of a car.
The demo covers the following:
- Multi-modal capabilities (text and vision for now, speech is coming to GPT-4o soon)
- Computer vision - extracting the features of a car from an uploaded photo
- Output using JSON Mode for consistency and accuracy
- Natural language text summarisation using different prompts
- (Optional) RAG based on the response from the DVLA VES API (third-party API, free API key available from GOV.UK on request)