Automate data extraction & analysis with intelligent document processing
summary
Broadridge automates extraction from SEC filings by building serverless, event-driven pipelines that process HTML filings at scale, turning millions of pages into structured outputs. Pipeline 1 uses Amazon Textract for tables (watch for API quota limits), Pipeline 2 uses Amazon Comprehend with custom entities for free‑text, and an A2I human review loop feeds corrections back into retraining. System replaces thousands of manual hours, covers ~5 million extracted data points across 300k meetings, and raises accuracy over time via continuous retraining.
AI-written summary · about
chapters (17)
- 0:00Intro
- 1:01The AWS ML stack
- 2:50Intelligent document processing (IDP) use case
- 3:49Document processing at scale is challenging
- 5:25Processing semistructured and unstructured documents
- 6:43Enhancing Al-enabled decisions with humans in the loop
- 7:46IDP with Amazon Al services
- 8:16Extract and analyze data from documents
- 9:39Broadridge (NYSE:BR) at a glance
- 10:36Document processing in the financial services industry
- 13:16IDP use case at Broadridge
- 14:56Approaches considered to automate document processing
- 17:17Selecting AWS Al-based approach
- 18:41Key architectural considerations
- 20:11Al approach/solution overview
- 25:31Key takeaways
- 30:48Learn ML with AWS Training and Certification
related talks
32mAutomate data extraction & analysis with intelligent document processing
30mIntelligent document processing: Extract data & insights from documents
35mAutomate your document pipeline with intelligent document processing
42mExtract information accurately from documents using computer vision
43mImplementing AI Document Processing at Scale: Syngenta
30mIntelligent document processing for the insurance industry
26mEmpower your business use cases
47mRevamping the Document Automation Workflow with Generative AI