Case Study: Leading Telecom Company achieves predictive prevention of API failures, $100K annual savings per API and 300% faster insights with Cognizant

Algorithms That Learn Are the Stars of the Self-Service Show

Cognizant partnered with a large telecom whose self-service system handles roughly 800,000 API calls a year. API failures hurt customer experience and drive up costs, so the company asked Cognizant to build predictive analytics that could proactively detect and prevent API failures in addition to its existing API management.

Cognizant developed a self-learning predictive platform that reads machine and application logs and applies random forests, neural networks, confusion matrices and advanced analytics (event classification, text mining, association rules, correlation mapping and time-series extrapolation) to model API behavior and failure patterns. The solution improved self-service success and customer satisfaction, cut management costs by more than $100,000 per API per year, increased the speed of insight generation by 300%, and reduced troubleshooting, infrastructure and deployment expenses.