Machine Learning

We have core competencies in Machine Learning and Big Data Analytics. We have developed applications in multiple domains for Global commodity players, from predicting prices using fundamental factors to predicting demand, supply, and crop yield based on user-driven parameters. We have also built and deployed counterparty credit default prediction solution, which helps our clients connect cash flow, counterparty margins, Counterparty risk and ratings to predict which counterparty is likely to default and thereby, take precautionary measures to avoid such scenarios. Some of our solutions also use inputs from IoT devices to predict the health of machines / pipelines / rigs.

Our Predictive Analytics applications are light, deployed on cloud and use in-memory calculations to provide real time output to the users. Some of our advanced, computationally intensive algorithms are designed to work on parallel computing frameworks, automatically sensing and utilizing all the available computing power to provide highly accurate results to users, fast. For processing large quantities of data, we use the latest Big-Data technologies like Spark with noSQL databases to create numerous clusters of large data banks and process each one separately using advanced machine learning algorithms. Our Big Data clients include a large US based Investment Management Firm, US State Government of New Hampshire, a large Data Management company in UK, among others.