Tulika Dutta

Data Scientist

Bengaluru, Karnataka, India10 yrs 6 mos experience

Most Likely To SwitchHighly Stable

Key Highlights

Expert in statistical modeling and machine learning.
Proven track record in optimizing retail operations.
Strong mentoring experience in data science.

Stackforce AI infers this person is a Data Science expert in the Retail industry with a focus on machine learning and optimization.

Contact

Skills

Core Skills

Data Science And Advanced AnalyticsMachine LearningDeep LearningStatistical ModelingOptimizationText Analytics

Other Skills

BigQueryCC++Clustering AlgorithmsConvolutional Neural Networks (CNN)Data AnalysisData ScienceDeep Neural Networks (DNN)GCPGitLarge Language Models (LLM)ManagementMicrosoft ExcelMicrosoft OfficeMicrosoft Word

About

Staff Data Scientist with 8 years of experience at Walmart, holding an M.Sc. in Applied Statistics and Informatics from IIT Bombay. Passionate about driving business value through statistical modeling, machine learning, and optimization. Core Competencies: - Python, SQL, R, Git, BigQuery, GCP. - Data Science and Advanced Analytics, Supervised and Unsupervised Learning, Textual and Image Data processing, Large Language Model (LLM), AI/ML model development/validation, Optimization, Statistical Modelling, Time series and forecasting, Clustering Algorithms - Recruiting and mentoring data scientists

Experience

Walmart global tech india

4 roles

Staff Data Scientist

Promoted

Nov 2025 – Present · 4 mos

PythonSQLRGitBigQueryGCP+4

Senior Data Scientist

Promoted

Jun 2021 – Present · 4 yrs 9 mos

Deep Neural Networks (DNN)Large Language Models (LLM)Machine LearningDeep Learning

Data Scientist

Aug 2019 – Jun 2021 · 1 yr 10 mos

Statistical Analyst

Aug 2017 – Jul 2019 · 1 yr 11 mos

Developed a Localized recommendation engine that suggests optimal assortment (F&C) i.e. set of products and number of products to keep for each Walmart store:
Created Customer Segments to identify similar purchase behavior from cash and card transactions.
Created Substitutable groups to find set of products where demand transference can be observed within the group.
Built non-parametric Choice and Demand Model at customer segment level for each substitution group within a category.
Constrained Optimization to find optimal set of products and facings using predicted demand subject to business and modular shelf dimension constraints.
Projected lift of 140 basis points on revenue.
Automated Item Linkage using Text and Image Attributes:
Walmart adds approximately 40 thousand new items each month and the time spent on manually assigning new items to appropriate line groups is substantial. To cater to the time criticality and huge pricing data, a solution has been implemented to automate the process of line assignment and whenever new items enters the eco-system it would be assigned to a line group based on the process output.
Implemented Random forest and Convolutional Neural Network / pretrained networks to learn text and image features to predict item classes.
Implemented SMOTE in K fold stratified random sampling techniques to handle class imbalance problem and Bayesian Model Combination as an ensemble method to combine two models.
Projected cost reduction of 4.5 Mn $.
Directed Diagnostics:
Identifying root causes for key business objectives (sales drop, excess inventory) across domains (assortment, pricing, supply chain), thereby helping merchants to make informed decision.
Created a framework to study the variation in sales volume for Walmart US stores, at an item market level, with respect to features from the domains of pricing, supply chain and assortment.

Statistical ModelingOptimizationClustering AlgorithmsRandom ForestConvolutional Neural Networks (CNN)

Placement cell, iit bombay

Company Coordinator

May 2016 – Apr 2017 · 11 mos

Facilitated placements for 1600+ graduating students by negotiating with national/ international firms andmanaging student – employer interactions and organized events (Pre-placement talks, aptitude tests) for graduating batch of students.

Syntel

Summer Intern

May 2016 – Jul 2016 · 2 mos · Mumbai Metropolitan Region

Developed a Text Analytics based Solution Platform using R to obtain value from social media data and to find product search relevance in e-commerce sites.
Its capabilities include a semi supervised learning algorithm using Sentiment Analysis, Text Summarization, Topic Modeling, Part-of-speech tagging, Word cloud generator, SVM, random decision forests and also variable selection techniques like PCA, latent semantic analysis.