➜ Client Name: Google (Team – Hardware Operations): ● Developed an Inventory optimization tool using demand/supply simulations, reducing distribution center inventory by $17M while maintaining 95% fill rates. The tool enables analysis of the impact of input changes on KPIs(Target Week of Supply), Service Level (SL), and costs across all Business Units ● Designed a custom ML Neural Network embedding model package using BigQuery SQL and Python to identify Proxy SKUs for New Product Introduction (NPI) with 97% accuracy. This tool aids planners in identifying proxies for difficult-to-source products ● Created a data-driven Portfolio optimization tool to optimize product launches, reducing costs by over $20M and streamlining operations. The tool analyzes factors including demand, supply chain constraints, and financial factors to optimize launch strategies ● Leading a remote team of 3 data scientists, developing roadmaps and validating models before production deployment in Vertex AI➜ Led a team of data scientists to develop an AI chatbot for a consumer hardware client using Google's PaLM 2 Bison model on Vertex AI. Optimized prompt engineering to improve the LLMs model efficiency and response accuracy, enhancing user experience

Partnered with cross-functional teams to implement MLOps pipeline for data validation and model monitoring for population health analytics projects, proactively addressing potential hurdles➜ Delivered $14 million in savings for healthcare clients by developing a Linear Regression model identifying at-risk high-cost Medicare members for targeted care management. Model improved accuracy by 7% over the CMS benchmark➜ Engineered advanced ML models to identify Medication Adherence members, leveraging Hadoop for data extraction. Achieving a 20% increase in medication adherence rates, delivering enhanced patient outcomes and cost-efficiency➜ Streamlined deployment of 4 ML models by transitioning codebase from SAS to PySpark, drastically reducing deployment time by 50% and enhancing efficiency

Spearheaded analytics teams for Search and Digital Marketing solutions; managed offshore staff, gathered business requirements and drove efficient workflows➜ Client Name: Microsoft Bing Search (Team – Marketing Analytics): ● Optimized sales strategy by designing and executing quasi-experimental testing and synthetic control methods to identify the most effective business model for sales representatives. Consistently earned the highest CSAT score ● Developed unsupervised K-Prototype models using ADLS data, significantly enhancing segmentation accuracy by 15% and boosting targeted sales performance by 10% ● Optimized Python program runtime from 8 hours to 12 minutes, drastically reducing manual work and improving operational efficiency ● Leveraged NLP topic modeling in Python to streamline issue resolution, resulting in a 30% reduction in issue resolution time ● Created executive Power BI dashboards and presented complex technical findings to senior leadership, enhancing project approval rates➜ Built a POC for the biggest bank in the USA using advanced time series modeling techniques to predict ATM cash demand with an error rate of less than 15%, resulting in optimized cash management and reduced operational costs

Led client success strategies in a fast-paced Ad-Tech startup, managing 4-6 accounts simultaneously while leading a 17-member team➜ Analyzed audience trends for blogger websites, implementing data-driven strategies that reduced marketing costs by 7% and enhanced campaign effectiveness➜ Led multicultural marketing campaign for Cricket Australia, resulting in increased coverage and convertibility, and generated AUD 30,000 profit

➜ Contributed to successful product releases through effective code testing and deployment processes➜ Collaborated within an agile framework to modify existing software functionalities and meet client requirements

➜ Worked as a liaison between the company and clients➜ Collected client requirements and worked with the technical team to deliver the output