Our client had thousands of contract documents scanned and stored as pdf documents. Manually scouring through these documents and searching for a specific clause or finding insight was becoming increasingly challenging day by day. Our challenge was managing hundreds of vendor contracts to understand and analyze clauses, renewal strategies, revenue, and legal implications. It required enormous amounts of text analysis.
Approach and Solution
- Our approach was to ingest these PDF documents into Azure and use an OCR service to convert them into text and perform text analytics.
- The PDF files were converted into PNG images through iText and stored in Azure blob storage. These images were then sent to Azure computer vision API for OCR. The extracted response from OCR API was pushed into Azure Search for indexing and search capability.
- The customer wanted specific clauses to be identified in each contract. This was done using NLP techniques to identify passages that were related to the clauses using text similarity matching. Also, a summarized version of every contract was made and indexed for a quick search.
- A specific feature built into the tool enabled our customer to select and extract specific regions of the contract, like a passage, tables, etc., and export the region for offline analysis. Predictive analytics based on specific costs was done to compare vendors over a period of time to understand which ones yielded cost benefits on repeated renewals. A web application based on processing contract details and providing analytics insights was built which served as a single go-to tool for all contract document-related requirements.
- Gain insights into your relationships with vendors, partners, and third parties.
- Improve efficiency in information search.
- Reduce organization risk, uncover hidden costs and opportunities.
- Analyze Vendor Non-Compliance and overpaying for unused product/services.
Our client had thousands of servers spread across the globe carrying internet traffic. The main challenge was managing these servers at optimal utilization.
and Wi-Fi Triangulation
Our client had thousands of contract documents scanned and stored as pdf documents. Manually scouring through these documents...
Automated image quality check
Our client receives a large number of crowd sourced images. They receive many pictures which are blurry and some of the images...