Kumud Joseph K. profile photo 1 on Tal

Kumud Joseph K.

Data Science/Machine Learning Leader

Bengaluru

15+ years of work experience

Narvar
Jan 20242 yrs 9 mosPresent

Bengaluru, Karnataka, India

Built IRIS, fraud/risk intelligence tool (https://corp.narvar.com/iris Developed Customer Identity Graph Scaled team from 0-to-x
Startups
Jun 20233 yrs 4 mosPresent

Mentored and served as an advisor to startups, helping them to adopt AI (including Gen AI)
Groupon
Oct 20193 yrs 7 mosApr 2023

United States/India

Merchant's demand forecasting - Design and managing team working on Groupon's merchant demand forecast to onboard best merchants. Created an impact of $1.7m in Gross Profit/year Merchant Attrition - Design & and managing team working on merchant attrition to help building retention strategies. Created an impact of $650k in Gross Revenue/year Product/Deal Bundling - Design & develop deals which are redeemed together frequently and create deal combos increasing the conversion rate to +2% Supply Density - Leading team in bridging supply gap corresponding to customers demand in major geographies with a potential impact of $3M/year Predicting Product Category - Using text data from Merchant Name & Deal Title SEO Keywords recommendation - Surfacing best keywords to help Agents write deal contents during a Deal setup. Resulted in yearly $750k incremental revenue in US Extreme Customer Experience flagging - Based on customer chat conversation, predicting if they have had any extreme experience. Resulted in yearly $445k savings worldwide + 4% increase in customer satisfaction Experimentation - Led Groupon's experimentation team and helped different teams run their A/B experiments smoothly and efficiently Machine Learning Platform - Oversaw and managed Groupon's first ML platform on Cloud(AWS)
Groupon
Apr 20163 yrs 6 mosSep 2019

India

Expired Deals Recommendation - Redirecting expired deals pages to closest live deal page. Resulted in yearly $379k impact + indirect improving deals SEO ranking Tour Deal Recommendation - For Upsell (related Tour deals) & Cross-sell (Goods deal) Agent Response Prediction/AI Chatbot Response - Predicting top response for a customer query/interaction with Groupon's Agents Search Engine via relevance - Improving Contact center search engine using relevance Refunds Prediction - Predicting customers who would come for refund in near-future Customer Segmentation - Segment customers based on wholistic view (value, profit, behavior, experience etc) so as to have differential treatments Campaigns propensity model - Identify right customer for upsell with focussed efforts Advanced Analysis of Refunds/Denials Impact
Groupon
Jan 20151 yr 3 mosMar 2016

India

Nearby Deal Recommendation -Algorithm to show best deals in customer's vicinity on Groupon Mobile App. Performance improvement of 15%(AU) and 58%(TH) in terms of GB reported. Additional GB of US$25k generated between Aug'15 to Oct'15 in SG Email Deal Relevance - Set of Algorithms and business rules, which facilitate relevant deals to users/subscribers through Emails. Test results showed 13% more conversion rate Dynamic Pricing recommendation -Automated Pricing correction algorithm based on performance of deals on Groupon site as well as competitor’s pricing. Testing on sample deal shows significant increase in GB by 68% and GR by 5% Product Bundling - Product bundling algorithm to decide which products (complementary/imperfectly competitive goods) should be sold together
Hewlett-Packard
Mar 20103 yrs 2 mosApr 2013

Evaluating Promotional/Marketing Effectiveness using Mix Modeling Developed a pricing/discount management tool for pricing decisions. Potential margin impact for this initiative is US$ 16-20 million for a year projects related to IPG-GSB Sales Team Sales Pipeline/Opportunities Win-Rate Analysis Account Sales analysis using RFM segmentation technique US Segmentation for optimum sales force allocation
Dec 20081 yr 2 mosJan 2010

Customer segmentation/value management for retention, development & win-back strategies Policy lapsation, Agent attrition/performace, Employee Attrition/Retention models for an Indian Insurance major Customer loyalty, perception & Customer Lifetime Value Models Building one entity-one record for customers, agents, policies

Kumud’s AI Stack

claude code
claude code
cursor
cursor
gemini
gemini
github copilot
github copilot
langchain
langchain

Education

Indian Institute Of Information Technology allahabad

BTech, Information Technology

Indian Institute Of Information Technology allahabad

20004 yrs2004

Join Kumud and 10k+ other professionals on

Tal Boss

Tal is the app where bosses meet engineers & builders directly, without middlemen

Download Tal Boss

Kumud Joseph K. on Tal

Headline

Data Science/Machine Learning Leader

About

Data Science, Machine Learning & applied AI leader with >20yrs of experience. Led global Data Science and Engineering teams and collaborated with Product and Business stakeholders in identifying opportunities and building DS/ML solutions around it. Expanded the team by 5X in just 18 months. Managed 21+ team of managers, data scientists, MLEs & analysts across different geographies. Designed robust & scalable solutions & collaborated with different tech teams in developing end-to-end products. Created roadmap for Data Science organization. Hired, mentored and built teams from scratch and managed projects generating $MM in revenue/year Areas:- Automation, Chatbot, Recommendation/Relevance, Demand Forecasting, A/B Experimentations, Dynamic Optimal Pricing, Product Bundling, Text Classification/NLP, Promotion Effectiveness & Planning, Sales & Pipeline Analytics, Attrition/Retention, Segmentation/Profiling, CRM, Marketing, Campaign Design & Management, Banking Frauds, Event Based Marketing, Cross-sell/Up Sell, Anti Money Laundering etc Industries:- E-Commerce, Contact Center, Product/Manufacturing, Banking, Insurance, Finance, Sales & Marketing Machine Learning: - GLMs, Tree Based Models(Bagging & Boosting), CF/Assoc, Clustering, SVMs, KNNs and many more. Deep Learning: - Transformers, Gen AI/LLMs, MLP/ANNs, RNNs Tools/Software:- Python/Pandas, Pytorch, Pyspark, R, SAS(Base & EM), SQL & sql-like(Hive), Databases(Teradata,Oracle), Hadoop (HDFS), Cloud (AWS)

Location

Bengaluru

AI stack

claude code

cursor

gemini

github copilot

langchain

Work experience

Director & Head of Machine Learning at Narvar, Jan 2024 - Present · 2 yrs 9 mos, Bengaluru, Karnataka, India, Built IRIS, fraud/risk intelligence tool (https://corp.narvar.com/iris Developed Customer Identity Graph Scaled team from 0-to-x

AI/ML Advisor at Startups, Jun 2023 - Present · 3 yrs 4 mos, Mentored and served as an advisor to startups, helping them to adopt AI (including Gen AI)

Senior Manager, Data Science at Groupon, Oct 2019 - Apr 2023 · 3 yrs 7 mos, United States/India, Merchant's demand forecasting - Design and managing team working on Groupon's merchant demand forecast to onboard best merchants. Created an impact of $1.7m in Gross Profit/year Merchant Attrition - Design & and managing team working on merchant attrition to help building retention strategies. Created an impact of $650k in Gross Revenue/year Product/Deal Bundling - Design & develop deals which are redeemed together frequently and create deal combos increasing the conversion rate to +2% Supply Density - Leading team in bridging supply gap corresponding to customers demand in major geographies with a potential impact of $3M/year Predicting Product Category - Using text data from Merchant Name & Deal Title SEO Keywords recommendation - Surfacing best keywords to help Agents write deal contents during a Deal setup. Resulted in yearly $750k incremental revenue in US Extreme Customer Experience flagging - Based on customer chat conversation, predicting if they have had any extreme experience. Resulted in yearly $445k savings worldwide + 4% increase in customer satisfaction Experimentation - Led Groupon's experimentation team and helped different teams run their A/B experiments smoothly and efficiently Machine Learning Platform - Oversaw and managed Groupon's first ML platform on Cloud(AWS)

Manager II, Data Sciences, Groupon Global Service Organisation(Robotics & Automation) at Groupon, Apr 2016 - Sep 2019 · 3 yrs 6 mos, India, Expired Deals Recommendation - Redirecting expired deals pages to closest live deal page. Resulted in yearly $379k impact + indirect improving deals SEO ranking Tour Deal Recommendation - For Upsell (related Tour deals) & Cross-sell (Goods deal) Agent Response Prediction/AI Chatbot Response - Predicting top response for a customer query/interaction with Groupon's Agents Search Engine via relevance - Improving Contact center search engine using relevance Refunds Prediction - Predicting customers who would come for refund in near-future Customer Segmentation - Segment customers based on wholistic view (value, profit, behavior, experience etc) so as to have differential treatments Campaigns propensity model - Identify right customer for upsell with focussed efforts Advanced Analysis of Refunds/Denials Impact

Manager I, Data Sciences, Groupon APAC at Groupon, Jan 2015 - Mar 2016 · 1 yr 3 mos, India, Nearby Deal Recommendation -Algorithm to show best deals in customer's vicinity on Groupon Mobile App. Performance improvement of 15%(AU) and 58%(TH) in terms of GB reported. Additional GB of US$25k generated between Aug'15 to Oct'15 in SG Email Deal Relevance - Set of Algorithms and business rules, which facilitate relevant deals to users/subscribers through Emails. Test results showed 13% more conversion rate Dynamic Pricing recommendation -Automated Pricing correction algorithm based on performance of deals on Groupon site as well as competitor’s pricing. Testing on sample deal shows significant increase in GB by 68% and GR by 5% Product Bundling - Product bundling algorithm to decide which products (complementary/imperfectly competitive goods) should be sold together

Analytics Manager, North America at Hewlett-Packard, Mar 2010 - Apr 2013 · 3 yrs 2 mos, Evaluating Promotional/Marketing Effectiveness using Mix Modeling Developed a pricing/discount management tool for pricing decisions. Potential margin impact for this initiative is US$ 16-20 million for a year projects related to IPG-GSB Sales Team Sales Pipeline/Opportunities Win-Rate Analysis Account Sales analysis using RFM segmentation technique US Segmentation for optimum sales force allocation

Analytics/Machine Learning Lead at Max New York Life/Satyam, Dec 2008 - Jan 2010 · 1 yr 2 mos, Customer segmentation/value management for retention, development & win-back strategies Policy lapsation, Agent attrition/performace, Employee Attrition/Retention models for an Indian Insurance major Customer loyalty, perception & Customer Lifetime Value Models Building one entity-one record for customers, agents, policies

Education

BTech, Information Technology, Indian Institute Of Information Technology allahabad