Near Real-time AI Solution for Automated Invoice and Order Data Extraction for an Apparel Manufacturing Company

Solutions

• All client data is stored in an Amazon S3 Data Lake, ensuring reliable and scalable storage. 
• Input documents undergo pre- and post-processing using AWS Lambda, applying image processing techniques like Gray Scaling and leveraging Amazon Textract for data extraction. 
• An AI-powered Optical Character Recognition engine is built on AWS serverless services, offering high scalability, reliability, and cost-effectiveness. 
• Amazon Textract extracts structured data, including forms and tables, by analyzing data relationships. 
• Amazon API Gateway facilitates seamless interaction between the Web UI and AWS system. 
• The results are post-processed and returned to the Web User Interface for easy access. 
• The Web UI, powered by React JS, incorporates real-time data modification capabilities and a user-friendly mapping feature. 
• Once data mapping is complete, the updated data is stored in Amazon DynamoDB through an internal API call. 
• The mapped data is further restructured and stored in the NoSQL database of Amazon DynamoDB, triggered by an AWS Lambda function via the Web UI. 
• The solution utilizes AWS services such as Amazon S3, Amazon Textract, AWS Lambda, Amazon DynamoDB, and Amazon API Gateway.