Cut through the fog of contractual information

Overview

Global enterprises need to navigate a complex landscape of contractual agreements in a multi-vendor ecosystem. The contracts differ in scope, are stored in silos, involve multiple stakeholders, and have legal implications for the underlying obligations. The scale and complexity of an enterprise’s repository of contracts demand a solution to manage, store, and interpret contractual information efficiently

Infosys Smart Contract Management solution capitalizes on artificial intelligence (AI) and machine learning (ML) to extract and distill intelligence from contractual documents in real time, allow access to stakeholders, and provide answers to specific queries.

Infosys Contract Analytics BizApps, built on our Nia AI platform, offers intuitive control by scanning documents using optical character recognition (OCR), direct feeds, or existing public repositories. It creates a configurable indexing capability and neural knowledge network from semantics embedded within contractual documents. Our solution provides a high fidelity dashboard and chatbot interfacing with underlying information.

Infosys Nia Bizapps solution analyzes documents and contracts using AI, ML, and OCR techniques to allow an intuitive interface. The engine ingests documents, links and stores them in a Hive data source. It uses key phrase extraction and undertakes supervised / unsupervised clustering to derive intent. The document parser and adapter uses complex rules to determine the best OCR algorithm to parse content. Infosys Nia selects one of 10 OCR extraction techniques to ensure a higher ‘hit ratio’ for extraction. It extracts images and text from portable document format (pdf) artifacts, other layered images, data from tables, removes horizontal and vertical lines, and automatically creates a relational structure between intent and response.

Infosys Nia stores predefined industry terms and builds a knowledge repository by adding new words to its domain dictionary. Our solution extracts data and content using tokenization, positional tagging, labeled dependency, named entity recognition extraction, and semantic similarity.

Several contracts have similar clauses, thresholds, and penalties. Infosys Nia applies the extracted rules, undertakes extraction, analyzes intent, and creates an internal learning map for correlating response to intent, based on the syntactic structure of a sentence.