How to Build Smart Applications Without a Machine Learning Background
summary
Practical roadmap for adding AI via managed AWS services using only HTTP calls and no ML expertise: use Amazon Rekognition Custom Labels to train on a few hundred images, use Amazon Textract (OCR++) to extract tables and forms (CRIF moves from ~60–65% to ~75–78% accuracy and cuts cost 25%), deploy Amazon Forecast and Personalize with your historical data, add Fraud Detector and Augmented AI for feedback loops, and run CodeGuru to find hot code paths (Amazon reports a 40% cost reduction on a Prime Day optimisation). Pay-as-you-go microservice integration is emphasised.
AI-written summary · about
chapters (14)
- 0:00Introduction and overview
- 3:14Why smart features matter
- 5:05New AI services in AWS
- 6:51Amazon CodeGuru features
- 12:00Amazon Fraud Detector explained
- 13:22Amazon Augmented AI overview
- 14:50Amazon Textract capabilities
- 18:55Amazon Rekognition features
- 21:27Custom Labels with Rekognition
- 24:22Amazon Forecast explained
- 26:29Amazon Personalize overview
- 35:21Building block approach
- 37:55Code demo and integration
- 40:24Resources and next steps
related talks
15mAI as a Service: Quick and Scalable Machine Learning POCs with No-Code/Low-Code Tools
36mWhat can you build with AI? No ML experience required
29mSolve common business problems with AI and ML
24mAccelerate your business with AWS AI Services
01h 41mGetting started with Machine Learning on AWS.
30mBring AutoML to all users across your business
05mMachine Learning for Developers
29mSolve common business problems with AI and ML