Mayur Kumar
Request intro

Mayur Kumar

Data Scientist

  • 39 proven skills
Request intro

Goes straight to Mayur. No recruiter spam.

Download résumé

Claims you can check

Each card points at something real — open the source and verify for yourself.

  • StatedRésumé

    GNN + LLM Integration for Stock Portfolio Optimization

    Designed and implemented a GraphSAGE-based GNN using PyTorch to model relationships between Nifty-50 stocks using financial and textual features. Integrated FinBERT for sentiment extraction and Mistral-7B for reasoning-based portfolio scoring. Built end-to-end ML pipeline including data preprocessing, feature engineering, model training with gradient optimization, and backtesting. Documented methodology as a formal M.Sc. thesis and IEEE/APA research article.
  • StatedRésumé

    Sentiment Classification with Deep Learning

    Processed 8,518 unstructured text records and built LSTM and CNN-based classifiers using PyTorch and TensorFlow. Applied NLP preprocessing including tokenization, vectorization, feature extraction, and evaluated models using standard classification metrics. Designed OOP-structured training pipelines for modular, reusable deep learning code.
  • StatedRésumé

    Weather ETL Pipeline

    Built an automated ETL pipeline for data extraction, transformation, and loading into PostgreSQL using Apache Airflow DAGs. Containerized services using Docker for reproducible deployment; implemented scheduling and monitoring for pipeline reliability.
  • StatedRésumé

    Civil Aviation Time Series Analysis

    Applied ARIMA, LSTM, and SatScan statistical models for trend analysis and pattern recognition on large aviation datasets. Co-authored a research paper presenting findings to academic reviewers; performed missing-data imputation and anomaly detection.
  • StatedRésumé

    SLA & Revenue Impact Analysis – Pizza Delivery Operations

  • StatedRésumé

    SENTIMENT ANALYSIS FOR PRODUCT REVIEWS

  • StatedRésumé

    Life Expectancy Prediction Using Machine Learning

  • StatedRésumé

    Weather ETL Pipeline using Apache Airflow

  • StatedRésumé

    SLA & Revenue Impact Analysis – Pizza Delivery Operations

  • StatedRésumé

    SENTIMENT ANALYSIS FOR PRODUCT REVIEWS

  • StatedRésumé

    Life Expectancy Prediction Using Machine Learning

  • StatedRésumé

    Weather ETL Pipeline using Apache Airflow

  • StatedRésumé

    GNN + LLM Portfolio Optimization

    Built a PyTorch pipeline combining GraphSAGE stock-relationship embeddings with FinBERT news sentiment, alongside a separate LLM-based scorer (Mistral-7B) that ranked stocks from prompted news summaries. Designed and ran backtests comparing four strategies.
  • StatedRésumé

    Sentiment Analysis with Deep Learning and MLOps

    Built and compared LSTM and CNN classifiers on 8,518 Flipkart reviews, selecting the final model by F1-score after feature engineering and hyperparameter tuning. Served the model through a Flask API on AWS EC2 with monitoring, and automated training and deployment with Prefect, tracking experiments and versions in MLflow.
  • StatedRésumé

    Time Series Forecasting for the Aviation Industry

    Prepared India civil-aviation financial data (handling autocorrelation, heteroscedasticity, non-stationarity) and compared ARMA, ARIMA, and LSTM forecasting models through structured evaluation. Ran SatScan spatial analysis to identify hotspots supporting infrastructure-planning recommendations; guided a team of three on model selection.
  • StatedRésumé

    Retail Sales Analytics Dashboard

    Cleaned and validated 10,000+ Blinkit sales records; engineered derived KPI metrics for segmentation and performance analysis. Ran exploratory analysis in SQL and Pandas to identify sales patterns, trends, and anomalies; built an interactive dashboard with drill-downs tracking Total Sales, Average Sales, Item Count, and Average Rating.
  • StatedRésumé

    GNN + LLM Integration for Stock Portfolio Optimization

    Designed and implemented a GraphSAGE-based GNN using PyTorch to model relationships between Nifty-50 stocks using financial and textual features. Integrated FinBERT for sentiment extraction and Mistral-7B for reasoning-based portfolio scoring. Built end-to-end ML pipeline including data preprocessing, feature engineering, model training with gradient optimization, and backtesting. Documented methodology as a formal M.Sc. thesis and IEEE/APA research article.
  • StatedRésumé

    Sentiment Classification with Deep Learning

    Processed 8,518 unstructured text records and built LSTM and CNN-based classifiers using PyTorch and TensorFlow. Applied NLP preprocessing including tokenization, vectorization, feature extraction, and evaluated models using standard classification metrics. Designed OOP-structured training pipelines for modular, reusable deep learning code.
  • StatedRésumé

    Weather ETL Pipeline

    Built an automated ETL pipeline for data extraction, transformation, and loading into PostgreSQL using Apache Airflow DAGs. Containerized services using Docker for reproducible deployment; implemented scheduling and monitoring for pipeline reliability.
  • StatedRésumé

    Civil Aviation Time Series Analysis

    Applied ARIMA, LSTM, and SatScan statistical models for trend analysis and pattern recognition on large aviation datasets. Co-authored a research paper presenting findings to academic reviewers; performed missing-data imputation and anomaly detection.
  • StatedRésumé

    SLA & Revenue Impact Analysis – Pizza Delivery Operations

    SLA & Revenue Impact Analysis – Pizza Delivery Operations
  • StatedRésumé

    SENTIMENT ANALYSIS FOR PRODUCT REVIEWS

    SENTIMENT ANALYSIS FOR PRODUCT REVIEWS
  • StatedRésumé

    Life Expectancy Prediction Using Machine Learning

    Life Expectancy Prediction Using Machine Learning
  • StatedRésumé

    Weather ETL Pipeline using Apache Airflow

    Weather ETL Pipeline using Apache Airflow
  • StatedRésumé

    Real-time Intern Demographic Analysis Data Recovery and Misconduct Identification

    Real-time Intern Demographic Analysis Data Recovery and Misconduct Identification
  • StatedRésumé

    Data-Driven Insights for Pizza Store Operations: Improving Delivery Performance and Revenue Management

    Data-Driven Insights for Pizza Store Operations: Improving Delivery Performance and Revenue Management
  • StatedRésumé

    Exploratory Data Analysis (EDA) of AMCAT Dataset and Hypothesis Testing: Investigating Salary Claims

    Exploratory Data Analysis (EDA) of AMCAT Dataset and Hypothesis Testing: Investigating Salary Claims
  • StatedRésumé

    Backend Refactoring and Bug Fixing for Note Taking Application Development and AWS Deployment of Regex Matching Web Application with Flask

    Backend Refactoring and Bug Fixing for Note Taking Application Development and AWS Deployment of Regex Matching Web Application with Flask
  • StatedRésumé

    Data Science Intern @ Innomatics Research Labs

    Collected and cleaned operational data from multiple sources, applying enrichment routines to produce analysis-ready datasets; ran hypothesis testing and SQL-driven analysis on the AMCAT dataset (10,000+ records) to validate assumptions. Built end-to-end ETL pipelines to transform and stage data. Deployed a Flask-based model service on AWS EC2 and added runtime monitoring to track model performance. Shared analysis results with the team to support delivery-performance and revenue-management decisions.
  • StatedRésumé

    Data Science Intern @ Innomatics Research Labs

    Performed EDA and hypothesis testing on structured datasets of 10,000+ records using Python (Pandas, NumPy, SciPy) to identify patterns and trends. Cleaned and transformed raw data through ETL processes; applied statistical techniques to support data-driven analysis. Built and deployed a Python-based regex matching application on AWS EC2 with REST API integration, improving automation and processing efficiency. Refactored Flask backend workflows, improving code structure, documentation, and application reliability.
  • StatedRésumé

    Intern @ Innomatics Research Labs

    Real-time Intern Demographic Analysis Data Recovery and Misconduct Identification
  • StatedRésumé

    Data Science Intern @ Innomatics Research Labs

    Performed EDA and hypothesis testing on structured datasets of 10,000+ records using Python (Pandas, NumPy, SciPy) to identify patterns and trends. Cleaned and transformed raw data through ETL processes; applied statistical techniques to support data-driven analysis. Built and deployed a Python-based regex matching application on AWS EC2 with REST API integration, improving automation and processing efficiency. Refactored Flask backend workflows, improving code structure, documentation, and application reliability.
  • StatedRésumé

    Intern @ Innomatics Research Labs

    Real-time Intern Demographic Analysis Data Recovery and Misconduct Identification
  • StatedRésumé

    Data Science Intern @ Innomatics Research Labs

    Real-time Intern Demographic Analysis Data Recovery and Misconduct Identification Data-Driven Insights for Pizza Store Operations: Improving Delivery Performance and Revenue Management Exploratory Data Analysis (EDA) of AMCAT Dataset and Hypothesis Testing: Investigating Salary Claims Backend Refactoring and Bug Fixing for Note Taking Application Development and AWS Deployment of Regex Matching Web Application with Flask
  • StatedRésumé

    2nd Rank - Fanalytics 2022

  • StatedRésumé

    Python For Everybody Certification Coursera Certification

  • StatedRésumé

    2nd Rank - Fanalytics 2022

  • StatedRésumé

    Python For Everybody Certification

    Coursera Certification
  • StatedRésumé

    Python For Everybody

    Completed certification on Coursera
  • StatedRésumé

    Second Rank, Fanalytics 2022

    Intra-Data Analysis Competition, Christ University

What Mayur can prove

Proven Unproven

See all skills →
INTERVIEW-VERIFIED CAPABILITY (PROVEN)SELF-DECLARED STACK
GRAPH NODES: 112 · CONNECTIVITY FACTOR: 0.35
Skill constellation: 112 skills, 0 interview-verified.AMAZON EC2AMAZON S3ANALYTICSAPACHE AIRFLOWAPPLIED MATHEMATICSARIMAARMAASTRO CLIAWS EC2BACK-END WEB DEVELOPMENTBASHBIG DATABUSINESS ANALYSISCIVIL AVIATIONCLASSIFICATIONCNNCODE REFACTORINGDATA ANALYSISDATA MODELINGDATA PIPELINESDATA PREPARATIONDATA PROCESSINGDATA RECOVERYDATA REPORTINGDATA SCIENCEDATA VISUALIZATIONDATABASE MANAGEMENT SYSTEM (DBMS)DEBUGGINGDEEP LEARNINGDELIVERY PERFORMANCEDEMOGRAPHIC ANALYSISDISTRIBUTION ANALYSISDOCKEREDAENGLISHETLEXPLORATORY DATA ANALYSISEXPONENTIAL SMOOTHINGFEATURE ENGINEERINGFINANCIAL METRICSFINBERTFLASKGITGNN / GRAPHSAGEGOOGLE COLABGRAPHSAGE GNNHINDIHTMLHYPOTHESIS TESTINGJAVASCRIPTJUPYTERJUPYTER NOTEBOOKKERASKPI REPORTINGLLMMACHINE LEARNINGMATPLOTLIBMICROSOFT EXCELMICROSOFT POWER BIMISTRAL-7BMLFLOWMODEL VALIDATIONMONGODBMULTILINGUAL NLPMYSQLNATURAL LANGUAGE PROCESSING (NLP)NEURAL LANGUAGE MODELSNLPNUMPYPANDASPIPELINE ENGINEERINGPOLICY DEVELOPMENTPOSTGRESQLPOWER BIPREDICTIVE ANALYTICSPREFECTPRODUCT ANALYSISPRODUCT VISUALIZATIONPYTHONPYTHON (PROGRAMMING LANGUAGE)PYTORCHRR (PROGRAMMING LANGUAGE)REGRESSIONREGRESSION TESTINGREST APIREST APISREVENUE GENERATIONSALES TRACKINGSATSCANSCIKIT-LEARNSCIPYSEABORNSENTIMENT ANALYSISSERVICE-LEVEL AGREEMENTS (SLA)SPATIAL ANALYSISSQLSTATISTICAL MODELINGSTATISTICSSTRATEGIC PLANNINGTABLEAUTENSORFLOWTEXT PREPROCESSINGTIME SERIES ANALYSISTIME SERIES FORECASTINGTOKENIZATIONTRANSFORMERSTREND FOLLOWING (TRADING)VS CODEWEB APPLICATION DEVELOPMENTWORD PROCESSINGLSTM
  • Amazon EC2: self-declared, 0 evidence items
  • Amazon S3: self-declared, 0 evidence items
  • Analytics: self-declared, 1 evidence items, linked to Project: Retail Sales Analytics Dashboard
  • Apache Airflow: self-declared, 2 evidence items, linked to Project: Weather ETL Pipeline, Project: Weather ETL Pipeline using Apache Airflow
  • Applied Mathematics: self-declared, 0 evidence items
  • ARIMA: self-declared, 2 evidence items, linked to Project: Civil Aviation Time Series Analysis, Project: Time Series Forecasting for the Aviation Industry
  • ARMA: self-declared, 1 evidence items, linked to Project: Time Series Forecasting for the Aviation Industry
  • Astro CLI: self-declared, 0 evidence items
  • AWS EC2: self-declared, 1 evidence items, linked to Project: Sentiment Analysis with Deep Learning and MLOps
  • Back-End Web Development: self-declared, 0 evidence items
  • Bash: self-declared, 0 evidence items
  • Big Data: self-declared, 0 evidence items
  • Business Analysis: self-declared, 0 evidence items
  • Civil Aviation: self-declared, 1 evidence items, linked to Project: Civil Aviation Time Series Analysis
  • Classification: self-declared, 1 evidence items, linked to Project: Sentiment Classification with Deep Learning
  • CNN: self-declared, 2 evidence items, linked to Project: Sentiment Classification with Deep Learning, Project: Sentiment Analysis with Deep Learning and MLOps
  • Code Refactoring: self-declared, 0 evidence items
  • Data Analysis: self-declared, 1 evidence items, linked to Project: Exploratory Data Analysis (EDA) of AMCAT Dataset and Hypothesis Testing: Investigating Salary Claims
  • Data Modeling: self-declared, 0 evidence items
  • Data Pipelines: self-declared, 0 evidence items
  • Data Preparation: self-declared, 0 evidence items
  • Data Processing: self-declared, 0 evidence items
  • Data Recovery: self-declared, 1 evidence items, linked to Project: Real-time Intern Demographic Analysis Data Recovery and Misconduct Identification
  • Data Reporting: self-declared, 0 evidence items
  • Data Science: self-declared, 0 evidence items
  • Data Visualization: self-declared, 0 evidence items
  • Database Management System (DBMS): self-declared, 0 evidence items
  • Debugging: self-declared, 0 evidence items
  • Deep Learning: self-declared, 2 evidence items, linked to Project: Sentiment Classification with Deep Learning, Project: Sentiment Analysis with Deep Learning and MLOps
  • Delivery Performance: self-declared, 1 evidence items, linked to Project: Data-Driven Insights for Pizza Store Operations: Improving Delivery Performance and Revenue Management
  • Demographic Analysis: self-declared, 1 evidence items, linked to Project: Real-time Intern Demographic Analysis Data Recovery and Misconduct Identification
  • Distribution Analysis: self-declared, 0 evidence items
  • Docker: self-declared, 1 evidence items, linked to Project: Weather ETL Pipeline
  • EDA: self-declared, 1 evidence items, linked to Project: Exploratory Data Analysis (EDA) of AMCAT Dataset and Hypothesis Testing: Investigating Salary Claims
  • English: self-declared, 0 evidence items
  • ETL: self-declared, 2 evidence items, linked to Project: Weather ETL Pipeline, Project: Weather ETL Pipeline using Apache Airflow
  • Exploratory Data Analysis: self-declared, 1 evidence items, linked to Project: Exploratory Data Analysis (EDA) of AMCAT Dataset and Hypothesis Testing: Investigating Salary Claims
  • Exponential Smoothing: self-declared, 0 evidence items
  • Feature Engineering: self-declared, 2 evidence items, linked to Project: GNN + LLM Integration for Stock Portfolio Optimization, Project: Sentiment Analysis with Deep Learning and MLOps
  • Financial Metrics: self-declared, 0 evidence items
  • FinBERT: self-declared, 2 evidence items, linked to Project: GNN + LLM Integration for Stock Portfolio Optimization, Project: GNN + LLM Portfolio Optimization
  • Flask: self-declared, 2 evidence items, linked to Project: Sentiment Analysis with Deep Learning and MLOps, Project: Backend Refactoring and Bug Fixing for Note Taking Application Development and AWS Deployment of Regex Matching Web Application with Flask
  • Git: self-declared, 0 evidence items
  • GNN / GraphSAGE: self-declared, 0 evidence items
  • Google Colab: self-declared, 0 evidence items
  • GraphSAGE GNN: self-declared, 0 evidence items
  • Hindi: self-declared, 0 evidence items
  • HTML: self-declared, 0 evidence items
  • Hypothesis Testing: self-declared, 1 evidence items, linked to Project: Exploratory Data Analysis (EDA) of AMCAT Dataset and Hypothesis Testing: Investigating Salary Claims
  • JavaScript: self-declared, 0 evidence items
  • Jupyter: self-declared, 0 evidence items
  • Jupyter Notebook: self-declared, 0 evidence items
  • Keras: self-declared, 0 evidence items
  • KPI Reporting: self-declared, 0 evidence items
  • LLM: self-declared, 2 evidence items, linked to Project: GNN + LLM Integration for Stock Portfolio Optimization, Project: GNN + LLM Portfolio Optimization
  • LSTM: self-declared, 4 evidence items, linked to Project: Sentiment Classification with Deep Learning, Project: Civil Aviation Time Series Analysis, Project: Sentiment Analysis with Deep Learning and MLOps, Project: Time Series Forecasting for the Aviation Industry
  • Machine Learning: self-declared, 1 evidence items, linked to Project: Life Expectancy Prediction Using Machine Learning
  • Matplotlib: self-declared, 0 evidence items
  • Microsoft Excel: self-declared, 0 evidence items
  • Microsoft Power BI: self-declared, 0 evidence items
  • Mistral-7B: self-declared, 2 evidence items, linked to Project: GNN + LLM Integration for Stock Portfolio Optimization, Project: GNN + LLM Portfolio Optimization
  • MLflow: self-declared, 1 evidence items, linked to Project: Sentiment Analysis with Deep Learning and MLOps
  • Model Validation: self-declared, 0 evidence items
  • MongoDB: self-declared, 0 evidence items
  • Multilingual NLP: self-declared, 0 evidence items
  • MySQL: self-declared, 0 evidence items
  • Natural Language Processing (NLP): self-declared, 0 evidence items
  • Neural Language Models: self-declared, 0 evidence items
  • NLP: self-declared, 2 evidence items, linked to Project: GNN + LLM Integration for Stock Portfolio Optimization, Project: Sentiment Classification with Deep Learning
  • NumPy: self-declared, 0 evidence items
  • Pandas: self-declared, 1 evidence items, linked to Project: Retail Sales Analytics Dashboard
  • Pipeline Engineering: self-declared, 0 evidence items
  • Policy Development: self-declared, 0 evidence items
  • PostgreSQL: self-declared, 1 evidence items, linked to Project: Weather ETL Pipeline
  • Power BI: self-declared, 0 evidence items
  • Predictive Analytics: self-declared, 0 evidence items
  • Prefect: self-declared, 1 evidence items, linked to Project: Sentiment Analysis with Deep Learning and MLOps
  • Product Analysis: self-declared, 0 evidence items
  • Product Visualization: self-declared, 0 evidence items
  • Python: self-declared, 0 evidence items
  • Python (Programming Language): self-declared, 0 evidence items
  • PyTorch: self-declared, 4 evidence items, linked to Project: GNN + LLM Integration for Stock Portfolio Optimization, Project: Sentiment Classification with Deep Learning, Project: GNN + LLM Portfolio Optimization, Project: Sentiment Analysis with Deep Learning and MLOps
  • R: self-declared, 0 evidence items
  • R (Programming Language): self-declared, 0 evidence items
  • Regression: self-declared, 0 evidence items
  • Regression Testing: self-declared, 0 evidence items
  • REST API: self-declared, 0 evidence items
  • REST APIs: self-declared, 0 evidence items
  • Revenue Generation: self-declared, 0 evidence items
  • Sales Tracking: self-declared, 0 evidence items
  • SatScan: self-declared, 2 evidence items, linked to Project: Civil Aviation Time Series Analysis, Project: Time Series Forecasting for the Aviation Industry
  • scikit-learn: self-declared, 0 evidence items
  • SciPy: self-declared, 0 evidence items
  • Seaborn: self-declared, 0 evidence items
  • Sentiment Analysis: self-declared, 2 evidence items, linked to Project: SENTIMENT ANALYSIS FOR PRODUCT REVIEWS, Project: Sentiment Analysis with Deep Learning and MLOps
  • Service-Level Agreements (SLA): self-declared, 0 evidence items
  • Spatial Analysis: self-declared, 1 evidence items, linked to Project: Time Series Forecasting for the Aviation Industry
  • SQL: self-declared, 1 evidence items, linked to Project: Retail Sales Analytics Dashboard
  • Statistical Modeling: self-declared, 0 evidence items
  • Statistics: self-declared, 0 evidence items
  • Strategic Planning: self-declared, 0 evidence items
  • Tableau: self-declared, 0 evidence items
  • TensorFlow: self-declared, 3 evidence items, linked to Project: Sentiment Classification with Deep Learning, Project: Sentiment Analysis with Deep Learning and MLOps, Project: Time Series Forecasting for the Aviation Industry
  • Text Preprocessing: self-declared, 0 evidence items
  • Time Series Analysis: self-declared, 1 evidence items, linked to Project: Civil Aviation Time Series Analysis
  • Time Series Forecasting: self-declared, 1 evidence items, linked to Project: Time Series Forecasting for the Aviation Industry
  • Tokenization: self-declared, 1 evidence items, linked to Project: Sentiment Classification with Deep Learning
  • Transformers: self-declared, 0 evidence items
  • Trend Following (Trading): self-declared, 0 evidence items
  • VS Code: self-declared, 0 evidence items
  • Web Application Development: self-declared, 0 evidence items
  • Word Processing: self-declared, 0 evidence items

Shipped work

See all work
No architecture case studies yet

Flagship systems this candidate architected will appear here once attached to the dossier. No project has been added yet.

Getting better, on the record

Proven skills over time — today marked in lime.

No scored sessions yet

This trajectory populates once the candidate completes scored interview sessions. No score has been recorded or estimated here.

Career chronology

Gaps shown honestly — only what's on the record.

Trajectory timeline

  1. Vellore Institute of Technology

    M.Sc.

    2024 — 2025

  2. Vellore Institute of Technology

    Master of Science - MS

    2024 — 2026

  3. Vellore Institute of Technology

    M.Sc.

    2024 — 2026

  4. Vellore Institute of Technology

    M.Sc.

    2024 — 2025

  5. Vellore Institute of Technology

    Master of Science

    2024 — 2026

  6. Vellore Institute of Technology

    M.Sc.

    2024 — 2026

  7. Vellore Institute of Technology

    M.Sc.

    2024 — 2026

  8. Data Science Intern

    Innomatics Research Labs

    2024-01-01 — 2024-03-01

  9. Data Science Intern

    Innomatics Research Labs

    2024-01-01 — 2024-03-01

  10. Intern

    Innomatics Research Labs

    2024-01-01 — 2024-03-01

  11. Data Science Intern

    Innomatics Research Labs

    2024-01-01 — 2024-03-01

  12. Intern

    Innomatics Research Labs

    2024-01-01 — 2024-03-01

  13. Data Science Intern

    Innomatics Research Labs

    2024-01-01 — 2024-03-01

  14. Python For Everybody

    2022-02-01

  15. 2nd Rank - Fanalytics 2022

    2022-01-01

  16. 2nd Rank - Fanalytics 2022

    2022-01-01

  17. Second Rank, Fanalytics 2022

    2022-01-01

  18. Christ (Deemed to be) University

    B.Sc.

    2021 — 2024

  19. Christ University

    Bachelor of Data Science and Artificial Intelligence honours

    2021 — 2024

  20. Christ University, Delhi NCR

    B.Sc.

    2021 — 2024

  21. Christ (Deemed to be) University

    B.Sc.

    2021 — 2024

  22. Christ University

    Bachelor

    2021 — 2024

  23. Christ University, Delhi NCR

    B.Sc.

    2021 — 2024

  24. Christ University, Delhi NCR

    B.Sc.

    2021 — 2024

Hiring for Data Scientist?

Request the full evidence dossier — verified interview sessions, cryptographically attested artifacts, and reference outcomes.

Request intro

Goes straight to Mayur. No recruiter spam.

Download résumé
Verification key —
Request intro