Automatically scale a serverless app with Amazon Textract & MongoDB
summary
Serverless pipeline that OCRs photos/PDFs with Amazon Textract, runs entity and key-phrase extraction with Amazon Comprehend, and stores native JSON in MongoDB Atlas so search and analytics happen in-place. It ingests the Enron corpus — over 500,000 emails and 34 TB of files — and claims to turn a four-year manual review into hours using S3, Lambda/Stitch functions, Atlas Search and Charts. Two engineers build the working demo in three weeks; async jobs and triggers handle scale.
AI-written summary · about
chapters (10)
- 0:00Introduction
- 1:08Enron
- 14:45MongoDB Atlas
- 16:36MongoDB Atlas Demo
- 20:01MongoDB Atlas Benefits
- 27:59Stitch Overview
- 31:18Atlas Search
- 34:37MongoDB Charts
- 37:12Recap
- 40:48Wrapup
related talks
13mAI Meets Serverless: Automating Receipt Intelligence with AWS Lambda & Textract
15mSave The World And Money with MongoDB Data Lakes
48mBuilding Serverless AI powered Applications on AWS
41mUnlocking Faster & Efficient Data Processing w/ Serverless
26mServerless Architecture for Product Defect Detection
38mBuilding Practical, Cost-Efficient GenAI Solutions Using Serverless
40mGoing serverless with MongoDB
46mHighly Scalable Image Storage with AWS Serverless