Cloud Powered AI / ML Use Cases
T 33
Customer Segmentation
Identifying customers in appropriate groups based on their similarity of transactions. It was developed with Clustering algorithm and helped running in focused promotion campaigns.
Tt 12
Customer Recommendation
Created a recommendation engine for customers using market basket analysis It helped identify spend behavior of customers which helped in improved cross-sell and up-sell
Tt 20
Sentiment Analysis
Unsupervised intent recognition to identify segments within the data. Very convenient for the law professionals to easily bifurcate judgements which are against the parties or favor to the parties.
Tt 14
Fraud Analysis
Improved fraud detection from existing rule based method to unsupervised deep learning based XGBoost algorithm.
Tt 17
Suspicious Transactions
Identification of suspicious transactions using rule based method, Spark structured streaming for Aadhaar Enabled Payment System.
T 88
Automation
Automatically identifying and classifying scanned documents using AWS TextRact
Tt 16
Recommendation Engine
Developing recommendation engine considering: similarity of content, browsing history, user behavior, interests, interactions, feedbacks, time spent etc.
Tt 15
ML Models Migration to Cloud
Migrating existing ML models from on-prem production to AWS Sagemaker and also build new models to leverage Credit card application data within big datalake on AWS.