Apr 2021 – Jan 2026 · 4 yrs 9 mos

• In my current role, I lead the Data Science team, focusing on advancing ML, OR, Generative AI adoption and driving optimization in manufacturing. I conduct bi-weekly knowledge-sharing forums to foster collaboration and continuous learning. My efforts in developing Text-to-SQL solutions have consistently achieved high validation accuracy, while I have also automated data cataloging processes to enhance data governance. Additionally, I have been instrumental in delivering predictive insights that have generated significant business impact across various projects.
• Achievements: I have successfully co-developed MILP scheduling algorithms that unlocked approximately $1M in potential value and enhanced ML models for lumber volume forecasting, resulting in a measurable business impact of $2–3M. Furthermore, I built advanced forecasting engines that delivered combined value of $7–9M, showcasing my ability to drive innovation and deliver results in the AI/ML domain.

Nov 2019 – Apr 2021 · 1 yr 5 mos

Nov 2018 – Nov 2019 · 1 yr · Bengaluru, Karnataka, India

• As the Principal Data Scientist, I led the Data Science team in developing and implementing advanced ML/DL algorithms for automated anomaly detection on large-scale time-series data. My role involved designing and deploying cutting-edge models that significantly enhanced the accuracy and reliability of anomaly identification for diverse customer use cases.
• Achievements: I successfully designed and deployed advanced anomaly detection models using LSTMs and ARIMA, which improved the overall performance of the system and provided valuable insights for our clients.

Jan 2018 – Nov 2018 · 10 mos · Bengaluru, Karnataka, India

• In this role, I developed and implemented advanced disaggregation algorithms to estimate appliance-level energy consumption using aggregated smart meter data. I also optimized core algorithms from MATLAB to Python, enhancing runtime efficiency and supporting large-scale customer implementations.
• Achievements: My work led to significant improvements in the accuracy of energy consumption estimates, contributing to better energy management solutions for our clients.

Mar 2016 – Jan 2018 · 1 yr 10 mos · Bangalore

• I managed a team of 6+ Data Scientists, overseeing the delivery of AI, ML, DL, and Big Data models for various projects. I drove innovation in ML automation and developed predictive maintenance models for an Oil & Gas client, as well as demand forecasting solutions for fuel consumption.
• Achievements: I successfully filed a patent for automatic hyper-parameter tuning of ML models and delivered impactful solutions that enhanced operational efficiency for our clients.

Aug 2014 – Mar 2016 · 1 yr 7 mos · Bangalore

Aug 2013 – Aug 2014 · 1 yr · Kolkata

• Responsible for Business promotion.Please find website for further details.
• http://www.myvideotalkindia.com/

Jun 2012 – Aug 2013 · 1 yr 2 mos · HSBC Electronic Data processing India Pvt. Ltd, Kolkata

• Marketing Analytics (Decision Sciences) – Providing end to end solutions that optimize marketing campaign design in terms of selection of right customers to be offered the right product at the right time, involved into business strategies like Customer Value Management, Contact Strategy, Shrink to Grow etc.
• Roles and responsibilities:
• Development and maintenance of all marketing propensity models across all savings, investment and FX products
• Development and automation (using SAS, advanced MS-Excel and VBA) of different customer and business dashboards to measure the business performance across various products and customer groups
• End to end project management starting from negotiating with stake holders in business about scope and timelines, arranging brainstorming sessions, doing rigorous hands-on analysis, obtaining project sign-off, project implementation and tracking results.
• Driving different cross team initiatives like Customer Value Management, Contact Strategy and Pricing synergy including representatives from model and product teams
• Key Projects:
• Customer Value Management: Bank has decided to prioritize the customers based on the value, instead of amount of deposits they hold, subsequently worked on integration of various key metrics to unify into single dimension in order to determine current and potential value, including Deposits (current & savings), Investments, Foreign exchange activities & Fee income, also key customer rates, margins and costs are applied to compute at individual customer level.
• Trail Commission Project: Due to the presence of multiple entries for re-investments, trail commission paid to wealth advisers inflated in AS 400 d. Situation has been highlighted to senior management and eliminated duplicates by different approach, which finally saved £ 40,000 per annum to bank.

Feb 2011 – May 2012 · 1 yr 3 mos · HSBC International, UK

• Key Projects:
• Fee sensitivity analysis: Bank has reduced the service fee on the customers, who ever maintains the minimum threshold balance, subsequently analyzed the impact on the attrition levels and suggested that, waiving fee income helped customers, having more balance than threshold but worsen the situation, who ever having less than threshold limit.
• Foreign exchange propensity model and potential analysis: Determined customer probability to perform Foreign exchange transaction using logistic regression framework. Significant variables have been identified and highlighted to business that customers holding multicurrency demand deposits with higher proportion of their wallet placed in demand deposits have high propensity, also performed analysis to determine the hidden potential income from the customers who do not have prior experience in transaction, which could be generated based on the existing significant characteristics of the customers who ever performed FX transactions, which has potential to increase in income by £ 5.8 m.
• Marketing effectiveness (Campaign analytics): Initiated the concept of marketing effectiveness to drive campaigns analytically by combining various key metrics across various channels, including actual deposits, wealth, FX and service levels in order to visualize overall performance instead of one by one in isolation.
• Office efficiency Analysis: Various dimensions and key metrics for all offices have been computed, which finally able to provide clear vision about all dimensions about offices.
• Attrition Forecast: Attrition analysis has been performed, which predicts the number of customers, who are going to attrite from the bank based on their tenure of relationship for both Premier and non premier segments.
• Office efficiency Analysis: Various dimensions and key metrics for all offices have been computed, which finally able to provide clear vision about all dimensions about offices.

Aug 2008 – Jan 2011 · 2 yrs 5 mos · HSBC Electronic Data processing India Pvt. Ltd, Kolkata

• Key Projects:
• Automation project: Yellow belt certified, Saved 0.25 FTE (equivalent to $ 10,500 / year) by automating the whole MI generation process, using windows batch programming, schedulers & lean methodology.
• Vista MI: MI provides key performance indicators related to Relationship managers, includes customers managing, funded customers, recruited customers, deposits managing, wealth holding and reviews performed, which enables the senior management to measure the overall performance of RM in one place.
• Customer central database preparation: By interacting with AS 400 and LPAR2 (Business objects), preparing customer central database by combining various demographics, product balance, contact preferences of customers on monthly basis, which is being used by team in Kolkata and Jersey for all analysis and MI purposes.
• Holistic Wealth management Universe: In order to improve the robustness in the preparation of MI wealth management has moved generation process from spreadsheet to data mart which directly uses the data from AS400.Worked with IT to generate various Business object reports including bonus statements for wealth advisers and automated bonus statement generation process, which process upfront and trail bonus to advisers.