● Developed AI products for Punchh Customer Engagement and Loyalty Platform for Customer Segmentation and Churn Prediction, Time Optimizations for Engagement Communications.● Facilitating end-to-end Machine Learning life cycle from data gathering, preprocessing, finding signals from noisy data, ML model training, performance evaluation, Model deployment, Data Pipelines for egressing data, and integration with the backend. The technology stack was Flask, Fast API, Airflow, Kubernetes, Python, and Data Science libraries like Scikit, XGB, TensorFlow, Pandas, and Databricks (Clusters, Workflows, Connect).

Role: ● Implemented projects in the financial services domain using advanced data modeling and analysis techniques to discover insights that will guide strategic decisions and uncover optimization opportunities on financial data.● Build, develop and maintain data models, reporting systems, automation, dashboards, and performance metrics that support key business decisions.● Performed analytics on relational datasets to identify root causes, patterns, trends, and recommend priorities. Communicated findings to the leadership team.Key Projects: ● Anti-Money Laundering Project: Used Anomaly Detection to flag suspicious transactions in known concerned corridors based on transactions data. Back-testing identified > $10MM of suspicious historical transactions.● Dashboards and Email Automation: Used Python’s Pandas, Requests, SMTP, and Flask APIs to automate reporting tasks and dashboards otherwise refreshed manually by the team. Reports range from single documents to full-stack dashboards. Trained offshore and onsite teams on the technology stack. Decreased turnaround time of reports by 10 times.

TCS A&I is a dedicated team of Analysts and Systems Engineers tasked with offering analytics based software solutions for businesses. I am privileged to deal with these huge datasets and developing end to end solutions for clients.- Responsible for delivering insights and performing predictive analysis in the space of Finance and Asset Management.- Involves heavy use of predictive algorithms ( such as Random Forest Regressor, SciKit Learn, TensorFlow, NLTK etc.,) and,- Integrating them with streaming analytics technologies like spark on HDInsights working on real time data coming in from APIs.- Implemented Kafka and Event Hubs based message queues to tackle live stream of data.- Also responsible for building Cloud Native Analytics Dashboards developed from scratch by developing RESTful APIs and dynamically updated web sockets based front end applications.

Worked on a ’ MOOC Course Recommendation ’ system given aUniversity curriculum the recommendation engine would suggestyou the best matching online courses or create a course packageto specifically cater to your course requirements.• Built the underlying machine learning model using text mining andcollaborative filtering. Was part of the team working on the fullstack Web Application for the same project.• Tools: Scikit-Learn, Jupyter, Numpy, Pandas, PHP, HTML, BootstrapCSS, ROutcome: Completed all Modules assigned to me on time

Worked as a trainee in Android Development stream focused on creating native mobile applications with complete utilisation of all the sensors and tech available on smartphones to cater to requirements in Public Service Domain. Extensive use of R language for data processing of Government Schemes results and Ministry datasets.Also Cloud technology including AWS EC2 to host REST API endpoints in PHP, MySQL and formatting the data on Smartphones using beautiful material design.Worked on a Dashboard app project required for completion of the programme (code on Github). With this projects bureaucracy and citizens can get better insights on success/impact of Government initiatives in a magazine format.