Gaurav Mahajan
Sr. Data Scientist / Associate Director · Mindtree Ltd
BangaloreWork Experience
Sr. Data Scientist / Associate Director
Mindtree Ltd2016 – Present
Achievements: • Business Impact: o GCPL: Increased Assortment and Sales by ~10% and ~3% respectively using Machine Learning techniques • Delivered 48X business impact (ROI) from Cross-sell recommendation model only • Delivered Strategic Insights with business impact which helped business optimize their salesforce team o Unilever - achieved 50% workforce efficiency through Automation project • Award: "Unstoppable" award for execution excellence in handling client expectations on target setting • New Business: Delivered POCs that resulted in New clients
Role and Responsibilities: • Lead Delivery and Innovation for "Assortment Prediction Recommendation Platform" for multiple clients including GCPL (Godrej Consumer Products Ltd), GSK (Glaxo Smithkline) and Johnsons & Johnsons. o Own the technical delivery/execution of APEX projects for FMCG clients and achieve primary objectives a). Oversee/Review the Recommendations b). Create Innovative Actionable Insights, hence the role comprises of both Data Scientist and Business Analyst. o Lead Client engagements (Strategic and Tactical level) including requirement gathering, fit gap analysis, model enhancements/improvements, recommendation performance reviews, root-cause analysis, insights and impact. o Built rapport and Trust with Clients -business and IT stakeholders including data scientists (mid/senior level) o Translating requirements into analytical specifications, implementations and designing solutions by following the standard guidelines o Drive Innovation/ Improvements to be brought in the platform - new KPIs design, Model experiments (new algorithms), Recommendation Optimizations design. o Creating & executing analytic solutions to improve & optimize business performance and minimize risk o Implementing & standardizing KPI metrics through comprehensive review & redesign resulting in better insights, consistent metrics and reduced service cost o Recognize opportunities and provide insights to enable the business to provide innovative solutions, creating customer value and growth o Determining different analytical approaches and provide a decisive view on practical uses; developing processes and trainings to institutionalize those methodologies o Build & test Hypothesis, using Statistical techniques, Root-cause-analysis and present results to the client. o Deliver Business impact from Recommendations to the clients, POCs and prospects on periodic basis o Responsible for Azure to AWS migration for a client o Key Decision maker in Big Data Migration, Models and Recommendations roll-out and analytic approach o Improve Prediction Accuracy at SKU-Store level using Machine Learning and Optimization: • Deliver 20 million recommendations every month at SKU-Store level on Azure Big Data • Built models to predict probability of selling SKU and its value/quantity in a store • Machine learning techniques: Logistic Regression, Collaborative filtering • Experiments on improving models using Neural Networks (Deep learning), XGBoost • Team management: o Allocate tasks, provide mentorship, facilitate training needs and regular feedback / appreciations o Work hands-on as well as drive the 10 member team (comprising of Data Scientist, Business Analyst, Data Analyst and Data Engineers) to achieve the project objectives • Planning Deliverables: o Create/follow-up on plans to successfully deliver multiple parallel projects on-time o Recognize potential concerns or opportunities and makes adjustments in order to respond to changing needs o Establishes internal checks in order to detect problems or issues and make necessary corrections o Manages multiple priorities and establishes timeframes and resources that are reasonable and attainable • Sales and Organization Support: o New Client opportunity - support the sales team in technical demo's and delivery of Proof-of-Value (POVs) o SOWs - provide effort estimates including Scope of work/Resource Loading & Planning and R&R o Interview Panellist to enhance Data Science team capacity building (primarily technical, few cultural) o Team Appraisals - provide feedback for direct reports and cross-functional team o Mentoring team including Junior Data scientists and Interns • Communication and Collaboration: o Client/Internal Stakeholder management - provide regular updates on the project to the senior stakeholders o Confident and fluent in giving presentation to top/middle management of client as well as internal and capable of responding to questions from the group o Collaborate with COE lead/team for new technology advancements incorporate in projects
Administering other Major Projects: • Lead Delivery of analytics/Insights projects for largest CPG retailer- Unilever (6 months): o New Product Launches Prediction: predicted success/ failure of New Product Introductions o Assortment Analytics - product and store recommendations using predictive models (Logistic regression) o Distributor / Store scorecard and segmentation using K-means/Decision tree/ Random forest o Country Diagnostics - provide 3600 insights related to Economy & Competition, Business Health, Customer development, Tactical execution and Business Initiatives. o New Retail Store Coverage Analytics - solution conceptualization to implementation. o Involved in Big Data Proposal for Unilever o Lead Technical enhancements - Data Model design o Responsible for Project P&L, Budgets, Investments and leadership discussions o Lead 8-10 member team of Statisticians, Team Leads, Business Analysts and Data Analysts • Increase Course adoption by successfully building Skills recommendation engine using Association rules- Yorbit (internal POC)
Data Scientist
Hewlett Packard- A&DM DataLabs2015 – 2016
Achievements: • Business Impact: E.On Churn Prediction USD 105Mn. revenue at risk delivered Shell Stock Forecasting delivered 99% accuracy Microsoft Skype Analytics reduced call drop from 10% to 2.5% • Award: received from NASSCOM Top 50 'Excellence in Analytics' 2015, for Real-time Analytics for HIMSS.
Role and Responsibilities: • Hands-on execution and Project lead for external clients in EMEA • Solution proposal, RFPs, Consulting, Project management, Deliverable, Collaterals preparation • Trained team, conducted Pseudo analytics projects • Published whitepaper (Linkden): Retail Marketer + Data science = Effective Marketing strategy • Major Projects delivered: o Microsoft Skype Analytics: Predictive analytics, EDA o Sponsorship Campaign Effectiveness for Shell: Predictive analytics, Text mining o Consumer review analytics for Q&A research (Marketing research & consulting, Netherlands): Predictive, consumer Sentiment/ Text mining, Descriptive and Guided analytics o Retail store Sales forecasting for Rossmann stores in Germany (Retail): Multivariate Regression o Call Volume forecasting for Accor group (Hospitality): Time-Series and Regression models o Customer Churn segmentation and Prediction for E.ON (Energy): Logistic regression, Classification Tree o Stock and Price Forecasting for Shell (O&G): Time-Series and Multivariate Regression models o Price Sensitivity Analysis for Automotive company: Regression model o Error detection for Carl Zeiss Vision (Manufacturing): Descriptive o Early Defect Detection for Automobiles: Descriptive o Real-time analytics for HIMSS (Healthcare): Descriptive • CRISP-DM methodology: to approach analytical solution • Models/ Algorithms used: Logistic/Multivariate Regression, Classification/Decision Tree (CART), Random Forest, KNN clustering, Association rule mining, RFM Scoring, Time-Series forecasting (ARIMA, Holt-Winters Exponential Smoothing, Moving Average), Naive Bayes, Principal Component Analysis. • Domains: Retail/CPG (core domain), Utilities, Automobiles, Healthcare, ITES, Hospitality, Oil & Gas • Statistical tools: R (primarily), SAS • Databases: Vertica, SQL • Visualization tools: Tibco Spotfire, Qlikview, Tableau
Program Manager - Analytics
Hewlett Packard- A&DM DataLabs2011 – 2015
Achievements: • Business Impact: Campaign Programs- delivered $7.7M incremental revenue. Sales Target Planning and Opportunity sizing: expected revenue increase for FY15- $10M approx. • Several Awards: for 3 consecutive years for developing retail solutions and generating business impact. • Clients recognitions: several appreciation emails from various stakeholders including increase ROI, optimize marketing spend, increased sales-force productivity, increased HP's revenues and share of wallet, opportunity assessment, improved competitiveness, retail best practices and improved processes.
Role and Responsibilities: • Led strategic Retail analytics solutions projects and consulting engagements for HP's top retail partners (Dixons and John Lewis) in EMEA-UK by working with HP UK consumer leadership team: • Trade Marketing Campaigns Planning and Effectiveness - objective: increase sales, optimize marketing spend and reduce contra spends for retail Demo day programs. Provided End-to-End campaign planning, tracking and business impact through campaign insights. Setup pilot vs. test stores (used Decision Tree model to segment stores; determined optimal Campaign duration, Stores and SKUs selection, focus stores and SKUs, expected sales, forecast ROI; sales targets per SKU, recommend in-store stock distribution, track performance and determine incremental sales and actual ROI. • Sales Target Planning and Opportunity Sizing - objective: increase potential sales opportunity size and determine optimal sales targets. Solution aims to plan sales targets by benchmarking stores, used clustering technique (decision tree model- to create and classify store clusters) to benchmark stores within a store cluster. Natural potential of the stores are determined by highlighting potential opportunity $ size. • Store Actionable Insights package - objective: improve sales-force productivity and margin. Built partner store management insights package, recommend SKUs where contra funds/ rebates could be saved; identify problem SKUs/Stores; Region/ Store/ SKUs performance using dashboards and scorecards. • Market Basket Analysis (POC) - recommended up/cross-sell products (using apriori algorithm) to RAMs. • Competition Landscape Analysis - provided insights and consulting on UK PC consumer retail market to drive premium SKUs break-through initiative driven by UK business team. • Product Assortment (POC) - cluster/classify stores and provided insights on SKU Range Gap Analysis (store vs. benchmark group) which helped to improve in-store product assortment. • Sales Forecast for Targeted Promotions (POC) - based on Customer demographics attributes (age, income, household size etc.) using Multiple regression model. • Time-series Sales Forecasting (POC) - using Regression/Holt Winter's/Exponential models. • Manage retail store database (Access/ SQL) which is used for various retail analytics projects. • Hands-on as well as managed the team size of 2-3 analyst.
Project Manager/BA
Manthan Software Service Pvt. Ltd2009 – 2011
Worked on ARC Business Intelligence projects with retail clients: Crocs Inc., USA (Footwear Retailer); S&R, Philippines (Hypermarket Retailer); Mydin Mohamed Holding, Malaysia (Hypermarket Retailer) and Puregold Price Club, Philippines (Hypermarket Retailer): • Project and engagement management across levels -CxOs, Department heads, managers. • Projects activities includes: strategy planning, analytical insights, consulting on campaigns, customer behaviour, competition and pricing areas to increase sales revenues and optimize marketing investments. • Provided insights on marketing programs, category and store operations for client's marketing, category and regional leadership team in development of strategic plan: o Customer Analytics and Insights: customer trends, customer behaviour, predicted sales demand from target customer segments (used historical customer demographics and behaviour trends to predict demand from key customer demographic attributes such income, age, household size etc.). Churn-Acquisition-Retention analysis, customer Loyalty analytics. o Campaign Insights: enabled business to increase sales by providing campaign insights on BTL campaigns (including Flyers, Direct Mail, Email Campaigns etc.) - identification of optimal promotion (used Multiple- regression model), promotion duration, products/ categories identification, target customer identification, tracking promotion performance and analyzed promotion impact through Pre-Post promotion sales increment lift and ROI. Identify customers segments (used RFM scoring) to maximize sales and optimize marketing spend. o Client has the objective to increase sales for their apparel product line from the limited budget. Implemented Market Basket analysis (association rule mining statistical technique), to identify cross-selling/ up-selling opportunity products such as which accessories go well with specific apparel products. This helped business in launching combo offers (buy-1-get-1 free or 50%off) and re-modify store shelf placement (in test stores) that helped business achieve 10% above their revenue objective. o Merchandise Assortment and Allocation- formed store clusters to allocate SKUs in right stores, category/sku range analysis, Season/Event based analysis, Stock out analysis. o Market Competition analytics - analysed landscape for products and suggested optimal price for the products under the promotion, using price elasticity (statistical technique). Analyzed Retailer vs. Market share & trends, Channel performance (store, online), Pricing comparison, Category mix comparison using external data like IRI, Neilson nitro, Neilson home scan. o New Store Location insights - identified Store (Sq.ft, size, Sales, location etc.) and Customer attributes (demographics- income, H/H size etc.; purchase behaviour) identify similar stores which helps to determine expected store Sales/ ROI. o Multi-Channel Analytics - Channel Profitability, Customer-Product-Channel Affinity
Sr. Consultant
HCL Technologies2008 – 2009
Worked with WMS Gaming Inc., Casino Retailer, USA, marketing team and provided analysis of the casino marketing programs, plan, support, analyze casino campaigns and promotions: • Assist in the development of short-term (eg. Happy hours promo offer) and long-term strategic marketing programs (eg. loyalty card reward program) in collaboration to improve revenues and profits. • Research and provide statistical information of new games in the market, player preference trends for new games in casino. Based on these statistics, provide insights on potential sales opportunity size for new games by analysing market, customer game usage patterns, analyse historical player/game usage trends. These insights helped the business in their strategic planning. • Used segmentation techniques (such as Demographic segmentation and RFM scoring) to segment customers/ players for a selected campaign and identify target customer groups, which helped business in promoting new games and player benefits.
Sr. Functional Consultant -Retail
Manthan Software Service Pvt. Ltd2006 – 2008
• Played key role in providing analytics, dashboards, scorecards and reports in ARC BI platform: o Store Operations Analytics - key analysis comprises Store Scorecard, Cluster based analysis, Store and Resource productivity analysis, Growth-Trend analysis, Like-for-Like Store analysis. o Merchandising and Inventory Analytics - Merchandise Plan Performance, Category scorecard, Category tactics including Assortment planning, OTB planning, Buying Plan, Allocation planning and Promotion planning; Pvt. Label vs. National Brand comparative analysis, Category range analysis, Season-to-Season analysis, Brand-style-color- size analysis, Stock ageing, Excess-short inventory, Stock Outs. o Customer Analytics - Demographic and purchase behaviour segmentation, Market Basket Analysis, Customer Churn-Acquisition-Retention, Loyalty based analysis, RFM Scoring, Campaign Analysis, Customer Concentration Analysis and Customer Purchase Behaviour Analysis. o Market Assessment (External Data Analytics) - Retailer vs. Competitor/ Market - Share & trend analysis, Channel Assessment, Retail format based analysis, Pricing Comparison between regions, Event comparisons, Category mix comparison, Private label/ National brand analysis, Promotion mix, Customer buying behaviour. This is achieved through data mapping external data like IRI, Neilson home scan data. o Vendor Performance Analysis - Vendor Performance/ Profitability, Vendor Scorecard, OTIF rate, Purchase Order based analysis, Rebate analysis o Promotions/ Marketing - performed analysis around Promotion Lift and ROI, Campaign/ mailers analysis, Markdown impact analysis, Discounts by attributes, Pricing Analysis. o Loss Prevention -Shrinkage impact, Exception analysis, Hot stores/ products/ customers. • Pre-sales - created and configured intuitive Role based Dashboards and Guided analytics. Conducted several demo's for fashion, food & grocery, speciality retail, mass merchandise retailers. • Documentation - provide functional documents to support Sales and Marketing teams. • Team Handling - train, mentor and manage team size of 2-3 associates in various projects
Business Development Manager
Targus India Private Limited2005 – 2006
Retail Store Manager
Sports Station (Nike & Levi's)2004 – 2005
Store Operations supervisor
IT TOOLS EXPERTISE2002 – 2004
• Statistics Tools: R (expertise), Python, SAS • Database: SQL, HP Vertica, MS Access • Visualization tools: Tibco Spotfire, Tableau, Qlikview • BI tools: SAP Business Objects, ORCA, ARC BI • Syndicated data tools: IDC, NPD, IFR, ACB, IPR • MS Office: Access, Excel, PowerPoint, Word
Education
MBA
Oxford Brookes University2003
MBA