Shivang Baijal

Product Manager

Delhi, India10 yrs 6 mos experience
Most Likely To SwitchAI Enabled

Key Highlights

  • Over 10 years of experience in Data Science.
  • Expert in developing AI recommendation systems.
  • Proven leadership in cross-functional team management.
Stackforce AI infers this person is a Fintech Data Scientist with extensive experience in machine learning and analytics.

Contact

Skills

Core Skills

Machine LearningData ScienceAnalytics

Other Skills

Fraud PreventionAI Recommendation SystemsDistributed Systems ArchitectureSQLNatural Language Processing (NLP)MLOpsPythonPySparkData AnalysisProblem SolvingTeam CultureAnalytical SkillsLeadershipTeam LeadershipNLP

About

Old School Data Scientist with more than 10 years of experience in Data Analytics and Data science domain. Organizations - Google, American Express, EXL Analytics, Siemens Programming Language - Python, R ETL - SQL, Netezza Scraping - Scrapy, Beautiful Soup Visualization - Tableau, Power BI, VBA Methodologies - Deep Learning (Tensorflow and Keras), LSTM, ANN, GAN, Regression, Decision Tree, KMean Clustering, Topic Modelling, NLP, SVM and ER Modelling (Database designing), Boosting and Bagging Domain - Investment and Portfolio Optimization, Digital Marketing Analytics, Healthcare(US), Finance, Churn Modelling and Insurance, Sales and Growth (Strategy), Hedge FundsOl

Experience

10 yrs 6 mos
Total Experience
2 yrs 1 mo
Average Tenure
3 yrs 2 mos
Current Experience

Mastercard

2 roles

Manager, Data Science

Promoted

Dec 2025Present · 6 mos

  • Currently leading the development and implementation of an advanced real-time AI recommendation system designed to address authorization decline surges across the Mastercard payment network. This initiative combines machine learning, distributed systems architecture, and fraud prevention expertise to optimize transaction approval rates while maintaining robust security standards.
  • Key Responsibilities & Achievements
  • Technical Leadership & Strategy
  • Spearheading the design and deployment of a real-time AI engine that analyzes transaction patterns and provides intelligent recommendations during decline surge events
  • Building and mentoring a cross-functional team of data scientists, ML engineers, and software developers to deliver scalable, production-grade solutions
  • Architecting low-latency recommendation systems capable of processing millions of transactions per second with sub-100ms response times
Machine LearningFraud PreventionAI Recommendation SystemsDistributed Systems ArchitectureData Science

Senior Data Scientist

Apr 2023Dec 2025 · 2 yrs 8 mos

  • 1. Acquirer/Issuer Recommendation Engine (Auth Fraud Recommender)
  • Analyzed Mastercard products including Tokenization, Auth Optimizer, and ABU to identify transaction approval rate issues.
  • Developed SQL-based rule engines to surface insights using business logic for anomaly detection.
  • Enhanced the system using a Cramer V-based correlation framework to identify variable combinations responsible for approval rate drops.
  • Translated variable pairs into human-readable recommendations, then fine-tuned a pre-trained LLM to automate this into Generative AI-powered suggestions for acquirers and issuers.
  • 2. Approval Rate Optimization via Payment Platform
  • Investigated historical transaction decline patterns to isolate data elements influencing approval outcomes.
  • Mapped variables controlled by Mastercard and merchants during transactions that led to avoidable declines.
  • Identified optimal data element modifications (e.g., formatting, compatibility) backed by evidence, achieving measurable approval rate uplift with minimal changes to the authorization request.
  • 3. Self-Learning Financial Assistant (LLM-based)
  • Fine-tuned a RAG-based LLM using PDFs, developer notes, and transaction datasets for RCA and on-demand financial analysis.
  • Enabled instruction-tuned querying over multiple datasets via a chatbot interface for merchants/acquirers.
  • Evaluated the system using Perplexity, ROUGE, and BLEU scores, and benchmarked on an in-house financial QA dataset.
  • Product actively used by customers to trace root causes of approval rate drops and suggest remediations across acquirer, merchant, or network layers.
SQLMachine LearningNatural Language Processing (NLP)AnalyticsMLOpsPython+2

Google

Senior Data Scientist

Jan 2023Apr 2023 · 3 mos · Gurugram, Haryana, India

  • Trust and Safety Analysis (Policies)
  • Data Augmentation and Seed Analysis: Performing Gap Analysis of pre-developed Machine Learning models for adult ads classification and filtering bad ad
  • Performing RCAs (root cause analysis) on patterns in bad ads and device strategies to avert them for optimal user experience
Problem SolvingTeam CultureAnalytical Skills

Plural investing llc

Principal Data Scientist || Strategy || Mixing AI with Human Intelligence

Nov 2021Dec 2022 · 1 yr 1 mo · Melbourne, Victoria, Australia

  • Reviews Analysis (NLP Modelling)
  • Develop Automated Web Scrapping Script to extract reviews for the target organizations, analyzed the key factors like quality, cost, selection etc. using supervised NLP (Boost, LSTM, Random Forest) and unsupervised (Topic Modelling) ML techniques, also performed Statistical Test to perform significance testing in Python
  • Stock Fundamentals Prediction (Quant Analyst)
  • Developed Time-Series Forecasting model for prediction of different Stock Fundamentals for Portfolio Optimization using LSTM, XGBoost, Random Forest, SVM and ANN and Recession Probability Model
  • Portfolio Optimization Model
  • Integration of Back testing with Machine Learning Technique to Optimize Portfolio Performance in Terms of CAGR, Fama French Factors,Sortino ratio etc.
  • Portfolio Optimization
  • Worked on Markowitz optimization to diversify portfolio selection
  • Used Cholesky decomposition to develop features for future covariance predictions enhancing portfolio diversification along with efficent frontiers to select the most optimal portfolios
  • Strategy
  • Developed automated analysis to augment decision making for portfolio diversification and optimization along with automated suggestions
  • Competitive and Market Analysis to understand the decision making for portfolio selections (Qualitative Research)
Machine LearningLeadershipData AnalysisProblem SolvingTeam LeadershipData Science

Upwork

Data Scientist

Oct 2020Nov 2021 · 1 yr 1 mo · Melbourne, Victoria, Australia

  • Top Rated Plus (Top 3%)
  • Worked on different project met different people across the globe. Some of the special projects on my account.
  • 1. Digital Marketing Analytics - Worked on forecasting and developing customer lifetime value
  • 2. ED Tech Instructor - Supported and worked with a prominent ed tech instructor
  • 3. Financial Analysis - Multiple projects on stock fundamental data using forecasting, machine learning and time series analysis
  • 4. SQL Developer - Developed ETL Processing Scripts and automated pipelines for automated data ingestion
  • 5. Interview Preparation Support - Prepared Multiple candidates for Interview (Data Science, Analyst and Engineers)
FreelancingMachine LearningNatural Language Processing (NLP)Global Client ManagementMicrosoft Power BIData Science

Mymoneykarma

Data Science Intern

Jul 2020Aug 2020 · 1 mo · Melbourne, Victoria, Australia

  • Customer Segmentation and Targeting for Campaign Effectiveness
  • ● Cohort Analysis using unsupervised machine learning techniques, classification of customers based on credit history and buying behavior to develop customer journey life cycle and develop a Markov chain model to estimate probabilities of next possible states of customer based on historical data points. Developed pickel files in Python to store model and implement in production
  • Technology Used - Python, SQL, PowerBI
  • Methodology Implemented – K-Mean clustering, Markov Chains and Sanky Diagrams
  • ● Probability distribution analysis - Extracted different sub-population within a wider population based on demographics and used the data to compute conditional and joint probabilities to rank the next most probable buy product to target each customer based on the ranked probability score.
Problem Solving

American express

Data Scientist, Risk Analyst

Jun 2019Feb 2020 · 8 mos · Gurgoan

  • Mapping unlinked card members
  • ● Developed Probabilistic Matching framework (Fuzzy matching) to enhance pre-defined rules of matching leveraging different attributes of customer to improve matching accuracy gauged using different metrics like False-Positives and True-Positives (underlinked and overlinked)
  • Product Analytics
  • ● Performing adhoc analysis, gap-analysis and impact analysis on different business problems for a leading new product launched by enterprise to unify the data across different markets and leverage features of different market to enhance decision making ability of the organization
  • ● Root cause analysis of issues like Data Ingestion and Data Linkage
  • ● Performing Exploratory Data Analysis on different attributes that can be added in our matching framework to enhance accuracy
AnalyticsLeadershipProblem SolvingTeam Culture

Exl

3 roles

Data Scientist (Assistant Manager)

Promoted

Jul 2018Jun 2019 · 11 mos

  • Data scientist - Worked on Churn Modelling and developed decision tree model to predict churn of customer base for US Based Life Insurance Client
  • Business Growth/Reporting - Developed different reports in Power BI (Python Based) and Tableau to pitch projects to new clients for the organization
  • AWS Developer - Worked on AWS environment to implement hive based queries on the datalake, to prepare the datasets required for Machine Learning model training
Problem SolvingTeam CultureAnalytical Skills

Senior Business Analyst (Data Science)

Promoted

Feb 2017Jul 2018 · 1 yr 5 mos

  • Marketing Analytics - Developed Logistic regression model and acquisition logic in SAS/SQL/Python, and delivered lead conversion reports in Excel VBA/ Tableau
  • Digital Image Feautre Extraction - Represented the company in Kaggel competition and developed algorithm to optimize model training and accuracy using texture analysis
  • DB Developer- Managed a whole project and developed data base and schema to report different quality and cost metrics for a healthcare provider using SQL and SSIS. Delivered final report in Power BI
Problem SolvingTeam CultureAnalytical Skills

Business Analyst (Data Science)

Aug 2015Jan 2017 · 1 yr 5 mos

Problem SolvingTeam Culture

Gannak analytics

Data Science Intern

Jun 2015Aug 2015 · 2 mos · Gurgaon, Haryana

  • Worked and developed Machine Learning models to forecast and optimize cost of raw material for an entire chain of raw material. Developed Dashboards using Power BI. Secondly, leverages SVM algorithm to implement logic for quality issue detection in a manufacturing unit.

Bharat heavy electricals limited

Intern

Jun 2014Jul 2014 · 1 mo · Bawana, New Delhi

  • Heat Recovery Steam Generation, Working of Power Plant, and Different kind of control valves in Power Plant

Netaji subhas institute of technology

Undergraduate Research work

Aug 2013Jun 2015 · 1 yr 10 mos · Dwarka, New Delhi

  • Instrumentation and Control Department (B.E)
  • Overall Percentage (74.0%)
  • Departmental Ranker 2 in Fifth Semester(85.26%)
  • Projects
  • 1. Project Under Ms.Smriti Srivastava(Dean UG)
  • Project on Human Identification, Activity Recognition (Bio-matrix)
  • Frame Extraction (from Video dataset), Background Removal, Data reduction (PCA, LDA), GAIT Energy Extraction, WEKA
  • 2. Project on Dicom image processing using MATLAB (Siemens)
  • Use of MATLAB programming to analyse the dicom file used in medical imaging, converting of DICOM into JPEG images converting JPEG into DICOM, Extracting dicom image information and editing the information of patient using MATLAB
  • 3. Project under Vijender Singh
  • Use of genetic algorithm and other optimization technique to determine the parameters of Different filter for comparison of their performances.
  • Publications
  • Under the Springer Publication
  • Advances in Signal Processing and Intelligent Recognition Systems
  • Volume 425 of the series Advances in Intelligent Systems and Computing pp 245-255
  • Performance Evaluation of S-Golay and MA Filter on the Basis of White and Flicker Noise
  • Main author of the above publication under the supervision of Shivangi Aggarwal and Asha Rani
  • The paper describes the comparison between the S-Golay and Moving Average Filter on basis of
  • SNR comparison using White and Pink(Flickr) Noise

Siemens healthcare

Marketing Intern

Jun 2013Jul 2013 · 1 mo · ITO,Delhi,India

  • AXA Basics, CT, Molecular Imaging, MR Basics, US, X-Ray, DICOM file
  • Project on Dicom image processing using MATLAB (Siemens)
  • Use of MATLAB programming to analyse the dicom file used in medical imaging, converting of DICOM into JPEG images converting JPEG into DICOM, Extracting dicom image information and editing the information of patient using MATLAB

Education

Monash University

Master 's of Data Science — Data Science

Jan 2020Jan 2021

Netaji Subhas Institute of Technology

Bachelor of Engineering (B.E.) — Instrumentation and Control engg.

Jan 2011Jan 2015

Department of Management Studies, IIT Delhi

Executive Management Programme

Jan 2018Jan 2019

C.R.P.F Public School

Central Board of Secondary Education — Science (non med)

Jan 1998Jan 2011

Stackforce found 100+ more professionals with Machine Learning & Data Science

Explore similar profiles based on matching skills and experience

Shivang Baijal - Product Manager | Stackforce