Viswanath Gangavaram profile photo 1 on Tal

Viswanath Gangavaram

Director of Data Science | AI Systems Leader building large-scale intelligent platforms across Ranking, Personalization, Ads Monetization, Logistics AI, Risk Intelligence, and GenAI ecosystems at scale

Greater Bengaluru Area

15+ years of work experience

Nykaa
Apr 20265 mosPresent

Building large-scale intelligent platforms across Ranking, Personalization, Monetization, Logistics AI, Risk Intelligence, and GenAI. Also, I’m serving as Product Leader for Nykaa’s Realtime ML Inference Engine, a self-serve deployment system for ML scientists.
Nykaa
Feb 20242 yrs 2 mosMar 2026

Bengaluru, Karnataka, India

I lead a 15-member team of Machine Learning Scientists at Nykaa Data Sciences, driving the Ads, Recommendations & Personalization (Reco & Pzn), and New Initiatives Pods. Over the past 1.9 years, we have executed 75+ A/B experiments delivering measurable business impact across the customer funnel. In Ads, we enabled foundational capabilities like Auto Bidding, Keyword Targeting, and Audience Targeting. We built and deployed ML models for Click Through Rate and Revenue per Click, combining them into a multi-objective ranking function CTR × (CPC+RPC) that optimized both organic and ad revenue—quadrupling ad revenue within a year. I also enabled manual/auto bidding, ad pacing options, and deployed multi-armed bandit models for homepage and listing banners. In Reco & Pzn, I launched multiple personalized widgets: last-minute add-ons, homepage LTR-based “Recommended for You,” brand/category widgets, and PDP modules like Visually Similar Items, Customer Also Bought, and Cart-Aware Pairings. I also built Nykaa’s first Learning-to-Rank powered vertical product listings on search, delivering >1% incremental Revenue per Visit. I spearheaded generative AI applications: Pair-It-With (cohesive outfit recommendations), Cart-Aware Pairings (checkout cross-sell), Trends-Based Widgets (influencer-driven catalog mapping), and Homepage Cross-Sell (personalized complements). Within New Initiatives, we deployed causal inference–based coupon/discount optimization, personalized MyOffers, Inventory Planning with forecasting methods, Expected Delivery Date prediction and initiated fraud detection with real-time ML inference. Additionally, I’m serving as Product Manager for Nykaa’s Realtime ML Inference Engine, a self-serve deployment system for ML scientists. Overall, I’ve scaled ad monetization, personalized customer journeys, and accelerated AI adoption across Nykaa’s ecosystem.
Swiggy
Apr 20221 yr 10 mosJan 2024

Bengaluru, Karnataka, India

Leading personalization chatter on the Restaurant’s menu and checkout screens. As part of this role, I am also a Tech solutioning mentor, and reviewer(Data Science) for ~8 data scientists to oversee the depth of their DS solutions. I am also tech solutioning mentor, and reviewer(DS) for inhouse Experimentation Platform) I own Data Science solutions for Recommended, Reorder, Cross-sell, Virtual Combos widgets in the Menu screen, ranking of categories, and ranking of items within those categories. I own personalization chatter on the checkout screen. I co-led the building of a general-purpose business-aware hyperlocal forecasting engine that can forecast at various spatial(geohash/zone/city/India/restaurant), and temporal granularity(next 10/30/60/120 mins, day, week, monthly) out of the box. Check the following detailed technical blog on my work on forecasting: https://bytes.swiggy.com/hyperlocal-forecasting-at-scale-the-swiggy-forecasting-platform-c07ecd5f5b86 I led Traffic & Acquisition Data Science(DS) chatter ( Sent >100 million PNs/day), you can find a detailed technical writeup on this work here: https://bytes.swiggy.com/smart-push-notifications-multi-armed-bandits-at-swiggy-part-4-f5698f2af0a6
Swiggy
Apr 20193 yrs 3 mosJun 2022

Greater Bengaluru Area

I co-led efforts on “general purpose anomaly detection, and explanation engine” which monitors business metrics at the hyper-local level, flags anomalies if any, and also gives first explanations on reasons for anomalies. Founding member of the Experimentation Council, which oversees all XPs that run at the organization. I led DS discounting efforts as part of this role. Employed experimental and observational Causal Inference to model the optimal discounting discussions. Please find my detailed technical blog on experimental causal inference: https://medium.com/swiggy-bytes/uplift-modeling-aka-heterogeneous-treatment-effects-evaluation-ad4e3fcf1d76 Other DS efforts I led Customer Life Time Value prediction, Churn Prediction.
Swiggy
Sep 20187 mosMar 2019

Bengaluru Area, India

I led the early discounting efforts as part of this role. Shown significant business impact within 3 months in this role.
[24]7
Apr 20153 yrs 6 mosSep 2018

Greater Bengaluru Area

- Conversational Agents, Sequence2SQL, Topic Modeling with text data - Web Site Personalization and Re-targeting Machine Learning models in multiple verticals - Measuring the combined "effectiveness of back end machine learning models and User Interfaces" through Online Controlled experiments - Digital Marketing
[24]7
Jul 20131 yr 9 mosMar 2015

Bengaluru, Karnataka, India

Working in R&D team. Primary job being to come up with novel machine learning algorithms.
[24]7
May 20121 yr 2 mosJun 2013

applications of machine learning algorithms in customer care and predictive sales. Applications of machine learning techniques for providing intuitive consumer experience.
Apr 20111 yrMar 2012

Post error correction mechanism by treating automatic speech recognizer (ASR) as a black box.
IIT Bombay
Jul 20109 mosMar 2011

We are working in Information Extraction field

Education

International Institute of Information Technology Bangalore

M.Tech, Databases and Information Systems

International Institute of Information Technology Bangalore

20082 yrs2010

Join Viswanath and 10k+ other professionals on

Tal Boss

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

Download Tal Boss

Viswanath Gangavaram on Tal

Headline

Director of Data Science | AI Systems Leader building large-scale intelligent platforms across Ranking, Personalization, Ads Monetization, Logistics AI, Risk Intelligence, and GenAI ecosystems at scale

About

A seasoned Data Science professional with over 15 years of experience applying machine learning on large-scale, high-impact use cases at Nykaa, Swiggy, [24]7.ai, HP Research Labs. On my name, I have 2 granted US patents, 2 research publications, and I published 6 technical blogs on Swiggy Bytes. I did an M.Tech in Computer Science with a specialization in Information Retrieval from IIIT-Bangalore, and after that, I was a Research Assistant to Prof. Soumen Chakrabarti at IIT-Bombay for 9 months. Check out my LinkedIn profile for more details on my background and experience. Github link: https://github.com/viswanath57 Medium link: https://medium.com/@visuthemoon

Location

Greater Bengaluru Area

Work experience

Director, Data Science at Nykaa, Apr 2026 - Present · 5 mos, Building large-scale intelligent platforms across Ranking, Personalization, Monetization, Logistics AI, Risk Intelligence, and GenAI. Also, I’m serving as Product Leader for Nykaa’s Realtime ML Inference Engine, a self-serve deployment system for ML scientists.

Senior Principal Machine Learning Scientist at Nykaa, Feb 2024 - Mar 2026 · 2 yrs 2 mos, Bengaluru, Karnataka, India, I lead a 15-member team of Machine Learning Scientists at Nykaa Data Sciences, driving the Ads, Recommendations & Personalization (Reco & Pzn), and New Initiatives Pods. Over the past 1.9 years, we have executed 75+ A/B experiments delivering measurable business impact across the customer funnel. In Ads, we enabled foundational capabilities like Auto Bidding, Keyword Targeting, and Audience Targeting. We built and deployed ML models for Click Through Rate and Revenue per Click, combining them into a multi-objective ranking function CTR × (CPC+RPC) that optimized both organic and ad revenue—quadrupling ad revenue within a year. I also enabled manual/auto bidding, ad pacing options, and deployed multi-armed bandit models for homepage and listing banners. In Reco & Pzn, I launched multiple personalized widgets: last-minute add-ons, homepage LTR-based “Recommended for You,” brand/category widgets, and PDP modules like Visually Similar Items, Customer Also Bought, and Cart-Aware Pairings. I also built Nykaa’s first Learning-to-Rank powered vertical product listings on search, delivering >1% incremental Revenue per Visit. I spearheaded generative AI applications: Pair-It-With (cohesive outfit recommendations), Cart-Aware Pairings (checkout cross-sell), Trends-Based Widgets (influencer-driven catalog mapping), and Homepage Cross-Sell (personalized complements). Within New Initiatives, we deployed causal inference–based coupon/discount optimization, personalized MyOffers, Inventory Planning with forecasting methods, Expected Delivery Date prediction and initiated fraud detection with real-time ML inference. Additionally, I’m serving as Product Manager for Nykaa’s Realtime ML Inference Engine, a self-serve deployment system for ML scientists. Overall, I’ve scaled ad monetization, personalized customer journeys, and accelerated AI adoption across Nykaa’s ecosystem.

Staff Data Scientist at Swiggy, Apr 2022 - Jan 2024 · 1 yr 10 mos, Bengaluru, Karnataka, India, Leading personalization chatter on the Restaurant’s menu and checkout screens. As part of this role, I am also a Tech solutioning mentor, and reviewer(Data Science) for ~8 data scientists to oversee the depth of their DS solutions. I am also tech solutioning mentor, and reviewer(DS) for inhouse Experimentation Platform) I own Data Science solutions for Recommended, Reorder, Cross-sell, Virtual Combos widgets in the Menu screen, ranking of categories, and ranking of items within those categories. I own personalization chatter on the checkout screen. I co-led the building of a general-purpose business-aware hyperlocal forecasting engine that can forecast at various spatial(geohash/zone/city/India/restaurant), and temporal granularity(next 10/30/60/120 mins, day, week, monthly) out of the box. Check the following detailed technical blog on my work on forecasting: https://bytes.swiggy.com/hyperlocal-forecasting-at-scale-the-swiggy-forecasting-platform-c07ecd5f5b86 I led Traffic & Acquisition Data Science(DS) chatter ( Sent >100 million PNs/day), you can find a detailed technical writeup on this work here: https://bytes.swiggy.com/smart-push-notifications-multi-armed-bandits-at-swiggy-part-4-f5698f2af0a6

Senior Data Scientist at Swiggy, Apr 2019 - Jun 2022 · 3 yrs 3 mos, Greater Bengaluru Area, I co-led efforts on “general purpose anomaly detection, and explanation engine” which monitors business metrics at the hyper-local level, flags anomalies if any, and also gives first explanations on reasons for anomalies. Founding member of the Experimentation Council, which oversees all XPs that run at the organization. I led DS discounting efforts as part of this role. Employed experimental and observational Causal Inference to model the optimal discounting discussions. Please find my detailed technical blog on experimental causal inference: https://medium.com/swiggy-bytes/uplift-modeling-aka-heterogeneous-treatment-effects-evaluation-ad4e3fcf1d76 Other DS efforts I led Customer Life Time Value prediction, Churn Prediction.

Data Scientist II at Swiggy, Sep 2018 - Mar 2019 · 7 mos, Bengaluru Area, India, I led the early discounting efforts as part of this role. Shown significant business impact within 3 months in this role.

Senior Data Scientist at [24]7, Apr 2015 - Sep 2018 · 3 yrs 6 mos, Greater Bengaluru Area, - Conversational Agents, Sequence2SQL, Topic Modeling with text data - Web Site Personalization and Re-targeting Machine Learning models in multiple verticals - Measuring the combined "effectiveness of back end machine learning models and User Interfaces" through Online Controlled experiments - Digital Marketing

Data Scientist at [24]7, Jul 2013 - Mar 2015 · 1 yr 9 mos, Bengaluru, Karnataka, India, Working in R&D team. Primary job being to come up with novel machine learning algorithms.

Analytical Consultant at [24]7, May 2012 - Jun 2013 · 1 yr 2 mos, applications of machine learning algorithms in customer care and predictive sales. Applications of machine learning techniques for providing intuitive consumer experience.

Research Intern at HP labs bangalore, india, Apr 2011 - Mar 2012 · 1 yr, Post error correction mechanism by treating automatic speech recognizer (ASR) as a black box.

Research Assistant to Prof. Soumen Chakrabarti at IIT Bombay, Jul 2010 - Mar 2011 · 9 mos, We are working in Information Extraction field

Education

M.Tech, Databases and Information Systems, International Institute of Information Technology Bangalore