Skip to content

Latest commit

 

History

History
36 lines (23 loc) · 742 Bytes

README.md

File metadata and controls

36 lines (23 loc) · 742 Bytes

Blume-hack

#Objective:

Build a platform to extract item & store details from billed receipt

#Tech stack

   1. AWS API Gateway
   2. Golang
   3. AWS Lambda
   4. AWS S3
   5. AWS Rekognition
   6. Kafka
   7. Elasticsearch
   8. Kubernetes

#Cloud Providers

   1. GPC
   2. AWS   
   
  P.S: We ran Kafka & Elasticsearch on GCP. AWS only provide 1 Core, 1 RAM for free tier :(

Design Diagram

Alt text

#Future Roadmap

  1. Improve data normalisation from cluttered image to text
  2. Rotation check of the images.
  3. Improve quality of image & then extract the text data
  4. Improve noise from retrieved textual data