Jun 2020 – May 2022 · 1 yr 11 mos · Stamford, Connecticut, United States

• Owned forecasting process for $2.5B business, responsible for statistical modeling, providing insights to sales and supply chain, gaining cross functional consensus, and conducting deep dive for S&OP meetings.
• Consistently achieved 85% forecast accuracy for Walmart, Target, Pharmacy and E-Comm channels.
• Reduced forecast bias to 2% by identifying forecast groups with 3 consecutive months of same bias.
• Improved summer in-stock for approximately 300 stores by recommending advance summer build.
• Improved time series accuracy by 24% by pipelining additional data from Walmart retail link system.
• Reduced 10hrs of weekly manual work by the team by creating Power BI dashboard for DPA tracking.
• Brought 14% improvement in supply chain efficiency by creating geographic forecast disaggregation.
• Saved approximately $6M worth of product from going into slob by identifying erroneous forecast.
• Drove approx. 1000bps improvement of downstream supply chain forecast at brand level.
• Created daily consumption vs. orders tracking dashboard to bring visibility to senior leadership.

Apr 2019 – May 2020 · 1 yr 1 mo · Seattle, WA, United States

• Led the contact forecasting team to forecast last mile delivery contacts in 10 countries at daily level granularity for 4 different contact channels. Responsible for managing forecast for $1B global network.
• Achieved overall 89% global forecast accuracy for 153 forecast group channel combinations level.
• Drove 500bps improvement in service level and occupancy metrics at customer service centers.
• Reduced manual reporting by 70% by creating automated daily reports in collaboration with BI team.
• Reduced recurring forecast misses by 60% through increased collaboration with stakeholders.
• Initiated fully automatic forecasting process for 4 countries by creating R based forecasting engine.
• Designed schema and created Redshift repository for storage and future intake of additional data.
• Established best practices by creating automated Tableau dashboards to reduce reporting errors.

Jan 2016 – Apr 2019 · 3 yrs 3 mos · Stamford, CT

• Unique role in the company performing weekly statistical demand forecasts and working alongside sales and supply chain team to perform ad hoc analytics for domestic and imported bottled water business.
• Managed forecasts for 12 accounts (7-Eleven, Publix, Jetro, Amazon etc.) worth approx. $1B business
• Delivered 72% accuracy on SAS time series forecast with 3-month lead time on more than 150 SKUs.
• Created Forecast Value Added dashboard in Power BI to identify 60 SKUs for automatic forecasting.
• Drove 20% improvement in accuracy/bias and supply chain performance for accounts covered.
• Directly contributed to over $2M in savings by creating R based price elasticity models for promotions.
• Identified 20 low performing SKUs by performing product segmentation using Nielsen and IRI data.
• Reduced over 100 manhours of weekly data entry in the organization by creating Excel VBA tool.
• Provided weekly drivers of forecast miss to company leadership as part of the S&OP process.

Jan 2015 – Dec 2015 · 11 mos · Bridgewater, NJ

• Worked on a client’s business problem of forecasting the optimum week in the growth cycle of pig hogs when they can be ready to sell. Created forecasting systems and reporting for client’s end customers.
• Brought 40% improvement on forecast accuracy by creating an R based forecasting engine for client.
• Managed a team of 3 offshore resources to maintain on time and efficient delivery of reports to client.
• Created approx. 20 interactive dashboards in SAP Lumira for customers to view on mobile and web.
• Improved customer retention 200% by creating ad-hoc reports and documentation for the business.

Sep 2014 – Dec 2014 · 3 mos · East Lansing, Michigan, United States

• Implemented Non Negative Matrix Factorization technique to perform text analytics on enterprise blog data, this project was funded by the National Science Foundation and broadly involved the following steps:
• 1. Cleaned more than 8GB of chatter data using of Postgresql from a spark data dump.
• 2. Performed Stemming and Sentiment Analysis on data using Python.
• 3. Usage of SAS Enterprise Miner to perform text clustering.
• 4. Generation of insights based on the results of the clustering model.

Jun 2014 – Aug 2014 · 2 mos · Greater Houston

• Internship as a Reliability Engineer at Baker Hughes Inc. provided exposure to the Oil and Gas industry and how it uses data for minimize cost due to tool failures and maximize operational profit margin.
• Primary project involved Root Cause Analysis of tool failures at rig sites in product families which are of concern to the company and discovering trends in the incidents taking place.
• The following are the key tasks which were performed by me at Baker Hughes:
• 1. Performed trend analysis on First Alert data for the Packers product line to identify problem tools and product families.
• 2. Worked on similar data for HSE(Health, Safety and Environment) team to predict potential HSE hazard incidents and gauge their business impact.
• 3. Helped identify product segments of concern across geo-markets on basis of profitability and Cost of failure.

Jan 2014 – May 2014 · 4 mos · East Lansing, Michigan, United States

• Served as a Teaching Assistant for a course in Discrete Mathematics (CSE260) for Spring '14 Semester, class of above 120 students.
• Duties during this period involved:
• 1. Grading of exam/homework and providing solutions to students
• 2. Office hours which were dedicated for doubt resolution of students.

Sep 2012 – Dec 2013 · 1 yr 3 mos · Gurugram, Haryana, India

• Part of market research team used survey data to perform descriptive and predictive analytics to solve client’s business problem. Responsible for data cleaning, statistical analysis, and questionnaire design.
• Conducted market segmentation for 12 countries using 2- Step and K-Means clustering techniques.
• Simplified customer behavior prediction by 2000% by developing a customer segment prediction tool.
• Generated insights from descriptive and exploratory analysis of Ad Eval and Brand health studies.