Better Understand Your Documents with AI
summary
Shows a concrete pipeline that turns varied files into queryable tables using OCR, AutoML Natural Language and BigQuery, with prebuilt processors created in two-click OCR processor style and endpoints for batch or streaming. Claims OCR covers over 200 languages and ~50 handwriting languages, AutoML trains in hours on hundreds of examples (not millions), and BigQuery runs on Colossus + Dremel with decoupled storage/compute and ultrafast network; result: structured data, scalable throughput, lower error rates.
AI-written summary · about
chapters (13)
- 0:00Introduction
- 1:20Agenda
- 2:25Customer expectations
- 3:44Challenges
- 8:19Document AI
- 9:18Document Understanding Flow
- 11:34Document AI Benefits
- 12:33Components of Document AI
- 16:27NLP
- 19:46Benefits of NLP
- 21:56BigQuery
- 23:26BigQuery Architecture
- 25:55Summary
related talks
01h 19mLearn to Unlock Document Intelligence with Open-Source AI
30mIntelligent document processing: Extract data & insights from documents
42mUsing Deep Learning to Understand Documents
31mIntelligent Documents Management System
35mAutomate your document pipeline with intelligent document processing
47mBuilding smarter apps with Machine Learning, from magic to reality
01h 00mComplex Data Ingestion with Open Source AI
17mAI/ML demo derby