Automated Purchase Order System

Challenges

Extracting the information from more than 400+ different and very complex layouts of purchase order.

Different languages of Purchase orders like Chinese, Vietnam and German.

Increasing dedicated resources to meet demand. Previous solution very expensive for OCR

Action taken

Configure rules based on each layout and transformation mapping to convert them to common flatten format. Step 1
Dynamically identify the language and extract the content based on the language Step 2
Business enrichment based on each customers of coats Step 3
Implemented In-house OCR engine with lowest Price Step 4

Actions Taken

  • Step 1

    Configure rules based on each layout and transformation mapping to convert them to common flatten format.

  • Step 2

    Dynamically identify the language and extract the content based on the language

  • Step 3

    Business enrichment based on each customers of coats

  • Step 4

    Implemented In-house OCR engine with lowest Price

The Problem

Extract thread purchase order and identify qty, colour and details for items with business based enriched contents.

The Solution

Automated extraction and enrichment of data at an affordable cost

The Result

95% Reduction in Manual Effort Average document processing time is 3 minutes.

82.5% Reduction in Cost, especially previously spent in purchase orders to flatten file conversion.