Primary designer and developer of end to end scalable but flexible pipeline facilitating no code model development/experimentation for over 200 models.Reduced the turn around time for deployment of all payment forecasting models significantly by leveraging model auto deployment framework for developing experimentation pipelines.

●   Part of a major firmwide payments forecast project development and responsible for driving the production ready deployment for over 30 models. This will benefit risk management, liquidity management as well as planning payment operation resources.●   Major contributor in taking Intelligent Selling solution live for Commercial Distribution Finance which is estimated to increase revenue by 7 million USD.●   Responsible for evaluating vendor AI/ML models for automated metadata… Show more ●   Part of a major firmwide payments forecast project development and responsible for driving the production ready deployment for over 30 models. This will benefit risk management, liquidity management as well as planning payment operation resources.●   Major contributor in taking Intelligent Selling solution live for Commercial Distribution Finance which is estimated to increase revenue by 7 million USD.●   Responsible for evaluating vendor AI/ML models for automated metadata discovery and providing detailed Model Development Document. ●   Lead Data Scientist in developing HR use case POC for Employee Attrition and Employee Displacement strategy. Show less

Facilitating Product placement strategy using Recommender Systems and Natural Language processing Techniques

Member of the Department Placement Committee of IISc, Department of Management studies. Responsible for organizing seminars from industry experts and co-ordinating with industry for placement, internships, and projects at the department.

Worked on multiple NLP based projects including multilevel text classification using ML algorithm, used state-of-the-art NLP transfer learning methods like BERT for text summarization and XLNET and MASS based natural language generation. Implemented latest XAI techniques like LIME, Anchor, and PDP on deep learning-based fraud detection model.

Ridership Prediction Model for BMTC: Involved in data collection and interaction with BMTC ITS to understand and set up the database for ticketing data and GPS data. Performed a comprehensive EDA for data set of over 200GB and presented the trend in ridership. Extracted reliability feature from GPS and ticketing data by analyzing the variability in travel time. Created a library in Python for cleaning, handling GPS data and integrating it with ticketing data to estimate travel time between… Show more Ridership Prediction Model for BMTC: Involved in data collection and interaction with BMTC ITS to understand and set up the database for ticketing data and GPS data. Performed a comprehensive EDA for data set of over 200GB and presented the trend in ridership. Extracted reliability feature from GPS and ticketing data by analyzing the variability in travel time. Created a library in Python for cleaning, handling GPS data and integrating it with ticketing data to estimate travel time between stops. Analyzed the effect of price variation on ridership and revenue in BMTC buses. Show less

Designed and automated tests for Virtualization and Performance features of the Solaris Operating System. Expanded tests to increase test coverage and support for the new generation of SPARC processor projects. Led test projects for performance counter and dynamic resource management features of Solaris O.S. Involved in designing and configuring new processors for Oracle Cloud Platform based on the previous customer issues, ensuring cross-platform migration, high availability, scalability and… Show more Designed and automated tests for Virtualization and Performance features of the Solaris Operating System. Expanded tests to increase test coverage and support for the new generation of SPARC processor projects. Led test projects for performance counter and dynamic resource management features of Solaris O.S. Involved in designing and configuring new processors for Oracle Cloud Platform based on the previous customer issues, ensuring cross-platform migration, high availability, scalability and virtualization Show less

Designed and developed low-level driver and API for monitoring performance using PMU in EMS ECU. Performed descriptive statistical analysis based on hardware level parameters of ECU.

Elected member of the Placement Committee of SJCE 2015-16 from the batch of 150+ students. Oversaw Placement training for 600+ students in collaboration with FACE academy. Verified and validated database of 1000+ students.