kunal kishore

AI Researcher

Bengaluru, Karnataka, India14 yrs 7 mos experience

Most Likely To SwitchHighly Stable

Key Highlights

Led a team of 17 Data Scientists at Walmart.
Filed 3 patents in data science innovations.
Expert in large scale information retrieval and ranking.

Stackforce AI infers this person is a Data Science leader in E-commerce and MarTech with extensive experience in machine learning and optimization.

Contact

Skills

Core Skills

Data ScienceMachine Learning

Other Skills

AnalyticsAuction designBusiness AnalysisData AnalysisData MiningGenerative AIGraph Representation LearningLLMLearning to RankLogistic RegressionMatlabMicrosoft ExcelMixed Integer Non Linear ProgrammingPredictive AnalyticsPython

About

Data Science Senior Manager with a demonstrated history of working in e-commerce, mobile marketing and advertising. Skilled in large scale Information Retrieval and ranking problems. Strong research professional with a Bachelor of Technology focused in Electronics engg. from Indian Institute of Technology, Kharagpur. My blog on data sciences: https://notesondatasc.wordpress.com/

Experience

14 yrs 7 mos

Total Experience

2 yrs 5 mos

Average Tenure

3 yrs 10 mos

Current Experience

Walmart global tech india

Senior Manager 2 Data Science

Aug 2022 – Present · 3 yrs 10 mos · Bengaluru, Karnataka, India

Leading a team of 17 Data Scientists and MLEs
Projects:
Yield management for entire Walmart display ads ($600M/year) by doing reserved price stochastic optimisation in first price auction setting
Automated campaign set up for advertiser's objective maximisation using large scale Mixed Integer Non Linear Programming
Guaranteed Delivery Allocation using Quadratic Programming optimisation at very very large scale (billion parameters)
Bid Landscape short term forecast (next 2 weeks) for display ads- using Log Reg model, Survival model
Bid landscape long term forecast (next 3 months) for display ads - incorporate booked demand, seasonality and time series demand forecast
Text generation using Generative AI - Display ad heading and subheading creation using state-of-the-art finetuned LLMs like LLAMA2, ORCA2
Image generation using Generative AI - Image generation for display ads using finetuned Stable Diffusion model
Filed 3 patents

SASMatlabMicrosoft ExcelData AnalysisPredictive AnalyticsStatistics+15

Udaan.com

Data Scientist (Principal)

Oct 2020 – Aug 2022 · 1 yr 10 mos · Bengaluru, Karnataka, India

Leading a team of 7 data scientist to solve medium scale data science problems in the discovery track at Udaan
Search Ranking - User response prediction, explore-exploit strategy, Learning to Rank
Search Relevance - Spell Correction, Store prediction, Unstructured Intent detection, Vocabulary Gap Matching, Extreme multi-class classification
Search Localisation - Translation and transliteration, Romanization, Catalog and search localisation
Query Autosuggest - Query autocompletion, query autosuggestion, ranking, Session based personalisation
Sponsored Search and Display Advertising - Auction design, User response prediction, explore-exploit strategy, Learning to Rank
Recommendation System - Personalisation, Repeat reco, Cart reco, Graph Representation Learning - multi-relation multi-entity graph
Audience/Persona Targeting - audience cohort for gmv maximization, marketing, CLM
Libraries/models/algorithms used: XMC, IndicTrans, lightGBM, Pytorch big graph, item KNN, symspell, Poisson distribution model.

SASMatlabMicrosoft ExcelData AnalysisPredictive AnalyticsStatistics+15

Inmobi

2 roles

Research Scientist & Manager

Promoted

Jul 2018 – Jun 2020 · 1 yr 11 mos

Image retrieval system
News title generation
Content Safety Filtering System
Lead Large Scale content Recommendation System
Large Scale ML solutions in MarTech (Marketing Technology) using CDP (customer data platform)
Teach Machine Learning to Software Engineers in internal Data Science BootCamps

Senior Research Analyst

Oct 2013 – Jun 2018 · 4 yrs 8 mos

Data Sciences:
Solve large scale Machine Learning problems related to ad-ranking in display advertising:
Click Through Rate prediction, Transaction per Click prediction
e-commerce recommender systems
Transaction Rate prediction
Market Place team:
Ad serve logic improvement for better margin and efficiency: auction theory/design, frequency capping, second price auction mechanism
Models
Used: Decision Tree, Logistic Regression, Image Retrieval, Clustering, Collaborative Filtering
Experimented with: Deep Neural Network for content recommendation system
Out of box Implementation: FM, UniLM for NLG, USE for NLP
Tools/Languages
Spark, Vowpal wabbit, Hive, Pig, sql, Python, Shell, R, Keras

Hsbc

Associate Analyst

Nov 2009 – Mar 2010 · 4 mos · Bangalore

Team: Default Analytics, Consumer and Mortgage Lending

Mu sigma inc.

Business Analyst

Jul 2009 – Oct 2009 · 3 mos · Bengaluru, Karnataka, India

- Built a churn prediction model using SAS for a technology firm

Hes-so haute école spécialisée de suisse occidentale

Summer Intern

May 2008 – Jul 2008 · 2 mos · Switzerland

Robotics.

Gopali youth welfare society

Vice President

Jul 2007 – Apr 2009 · 1 yr 9 mos

Founded "Jagriti Vidya Mandir", a free of cost English Medium School in the village of Gopali, 5 km from IIT Kharagpur campus. Worked on expansion of the school by organizing various events and thus brought several local industries and organizations in touch. Led a team of 15-20 members.
Link:
http://www.gyws.org
https://en.wikipedia.org/wiki/Indian_Institute_of_Technology_Kharagpur#Gopali_Youth_Welfare_Society