Computer Vision in Azure
summary
Three production demos show concrete outcomes: invoice processing uses Azure AI Document Intelligence plus a Python Azure Function pipeline that polls Gmail every five minutes, runs eocr OCR, renames PDFs by extracted invoice number and yields ~90% confidence and far fewer manual errors; a mobile waste app uses Vision Studio models, blob + SQL storage and Azure Functions with Azure OpenAI to estimate expirations and generate recipes, improving sorting and reducing food waste; flow‑meter reading runs a Streamlit frontend calling Azure OpenAI GPT‑4o multimodal to return meter values and tracks token costs per image.
AI-written summary · about
chapters (9)
- 0:00Computer Vision in Azure
- 18:09Agenda
- 24:18Invoice processing with Azure AI
- 44:41Python code for invoice automation
- 54:25Waste management with Azure AI Vision
- 1:04:02Waste recognition app architecture
- 1:12:40Reading flow meters with Azure OpenAI
- 1:21:10Flow meter reading demo
- 1:31:37Q&A
related talks
46mBuild apps with state-of-the-art computer vision
30mCreate Smart Angular Apps Using Azure Computer Vision
29mApplied Custom Vision
46mBuilding computer vision applications Q&A
01h 00mLearn Live: Get started with AI on Azure
50mScience Art in the Cloud: How Microsoft Azure helped me to Express Myself
01h 10mSolutions prototyping using Azure AI Studio
01h 02mThe Modern Cloud