AI & Data Science leader with a PhD in Applied Economics and 18+ years of experience in building and deploying business solutions leveraging Analytics, Data Science, and AI across diverse industries. Currently Leading the Data Science CoEin Accenture, consisting of around 100 data scientists, Data Engineers, AI engineers and BI developers with a focus on executing ML/AI & GenAI based POCs, client projects, developing AI assets cutting across multiple domains and industries. Seeking a leadership role in AI & data science to drive digital transformation initiatives and foster innovation using AI and ML. Holds B1/B2 visa, valid upto Aug 2029. Willing to relocate for suitable roles.
o As DS CoE lead, managing around 100 Data scientists, AI & Data Engineers and BI developers of various levels, executing Best-in class AI and data science work across industries and clients
o Manages a P&L of ~USD 6.5 M revenue and with a business value impact of ~USD 250M
o Drives the Model Audit/ review governance process for DS delivery teams
o Leads Knowledge Management & innovation initiatives – Filed 4 patents & 3 of them got granted, Organised Hackathon & Whitepapers contests.
o Plans and spearheads New Capability Development for DS CoE. E.g, GenAI, Simulation Modeling and Optimisation tools & techniques
o Manages the hiring & training for DS CoE
o Works with Presales & BD on RFI/RFP, client pitching, Solutioning & Statement of Work
Advanced technical competency in AI / ML Algorithms, Generative AI and Conversational AI, Deep learning, Neural Networks, Simulation modeling, Predictive modelling, Forecasting, Optimization, NLP & Text analytics.
Published 5 research papers, 4 conference papers & Edited 2 Corporate reference books in a wide range of macro and applied economic topics in national & International journals.
Basic SAS by SAS institute
Level2 mortgage analytics expert by CoreLogic
AWS Elevate by AWS inc.
· GenAI based Real time event and Risk detection using public data:
Leveraged LLMs and multi-modal foundation models, to generate real-time alerts and insights related to critical events and risks such as natural disasters, financial frauds, cyber security incidents, violences, etc for corporates, Govt organizations and news agencies. The solution enabled the customers to determine the impact of the risk and respond quickly and effectively.
· Procurement (PR to PO) Process automation using AI/GenAI tools and methods;
In an attempt to automate the PR to PO validation as part of the touchless procurement process goal, used information extraction and matching of text in the PR documents using GenAI / NLP methods and built a rule engine which automated the end-to-end PR to PO process , which significantly reduced the involvement of human agents in the process.
· Advanced Inventory optimization for supply chain & procurement organizations:
A combination of heuristic and statistical approach to estimate optimal safety stock to be maintained for goods whose demand are highly intermittent and sparse for a oil & natural gas company. Used Bootstrapping technique to smoothen the series and to estimate the min level of stock to be maintained. Used heuristic algorithms to calculate optimal demand variability and lead time variability which were eventually used for estimating optimal safety stock for individual SKUs belonging to Spares & consumables category. The solution resulted in a 35M$ inventory cost reduction for the client business.
· Vessel Schedule Optimization for Oil & Gas company:
Built a heuristic solution framework to determine the least cost combination of vessel routing & scheduling in servicing oil wells for an Oil & Gas company in south east Asia.The algorithm takes in to account the constraints in terms of Vessel type, time , spread type, etc. while determining the optimal route plan, the cost per route and eventually the least cost route which resulted in a savings of 25MR for the business.
· Conversational AI for an Analytical Ops Tool: Building a conversational AI tool (virtual Assistant/Chatbot) using LLMs and lang chain models which is to be deployed on AWS cloud platform.
· Building Digital twins for business Ops using Simulation models in AnyLogic:
Developing Digital twins of business Ops processes and using them to understand the impact of digital & AI transformations through simulation & optimization models. Mapped the processes such as Order To Cash, Talent Acquisition, warranty Claims in terms of As-is and To- be scenarios and measured the impact on process metrics and business outcomes using multi method simulation models using AnyLogic tool.
· Default & prepayment models in Mortgage Risk Management: working with a renowned US mortgage service provider, designed and built a major Risk prediction tool which included development, testing and validation of default and prepayment models using AI/ML algorithms, which were rolled out as a financial product in the US mortgage industry.
· Residential Property valuation models: Used a combination of Bray-Curtis distance measures and similarity ratios to build an in-house algorithm that quickly assesses the suitability of comparable properties in estimating the value of a given property.
· Scorecard Development – Acquisition models: Developed application level scorecard for auto loans using logistic regression technique in SAS environment.
· Call Volume Forecasting & Work Force Manager for a healthcare business: Using Ensemble forecasting system and advanced Optimization algorithms developed a Work Force Manager (WFM) for managing the workflow of customer care center for healthcare businesses. Inbound call volumes were forecasted using a ensemble forecasting tool consisting of 20+ algorithms including neural networks, Random Forest, Bayesian and Fourier transformations. Based on the forecasted call volumes, workforce capacity planning and agent scheduling workflows were determined using Erlong C function and LPP optimization models. This resulted in around 20% reduction in staff requirements.
· Sales enablement forecasting models for a Healthcare business: Used Two-staged segmentation & logistic regression models to predict the conversion potential of a contract in the healthcare equipments industry. Variables related to the deal and customer demographics were used for creating CHAID based segments which were in turn used for building logistic regression prediction models.
· Smart AI Ops solutions and Win probability predictor for a consulting major:
Built a logistic regression model to predict win probability of sales pipeline and used simulation to make more informed BD decisions. Applied advanced ML & AI algorithms including deep learning (ANN) - as well as sophisticated mathematics to validate assumptions. Through the implementation of the model, 10-12% of Total BD Spent was made available to be spent on better selected Opportunities
Python, SAS, R
Sports, Travel and Fitness