Revamping the Document Automation Workflow with Generative AI
summary
Azure AI Document Intelligence demonstrates a new custom generative model that produces grounded extractions with word‑level spans and confidence scores, supports zero‑shot extraction and schema validation, and exposes a buildable inference endpoint; it combines layout, markdown paragraph/table outputs and Query fields to normalize keys. PwC surfaces this in a RAG pipeline powering ChatPwC for 200,000 users with under 2% failure rate after one year in production.
AI-written summary · about
chapters (12)
- 0:00Document automation with GenAI
- 1:17GenAI transforms document processing
- 3:47Core workflow patterns
- 7:17Azure OpenAI challenges in RPA
- 9:02Custom models for extraction
- 14:41Handling complex documents
- 21:50Custom model availability & next steps
- 22:33ChatPwC with Azure Doc Intelligence
- 25:17RAG architecture & Doc Intelligence role
- 34:03Advanced Azure Doc Intelligence
- 44:03Custom extraction for audits
- 45:47Q&A
related talks
43mA journey from static documentation to generative AI
50mAzure AI, Power new possibilities for every organization
48mGetting started with generative AI using Azure OpenAI Service
46mAI innovation in the Microsoft Power Platform
20mWhy Intelligent Document Processing is AI’s Sweet Spot
24mInfuse Generative AI in your apps using
22mLearning to Apply Generative AI at Enterprise Level
15mLatest Generative AI products and solutions on