KISHAN SHELADIYA

ML + Data Engineer

Hi, I'm Kishan Sheladiya.
I build ML systems and data-driven products.

Data Science & ML Engineer with hands-on experience building ML models, data pipelines, and analytical solutions. Currently completing an M.Sc. in Mathematical Modelling, Simulation & Optimization at Universität Koblenz (expected Sep. 2026). Skilled in Python, SQL, PyTorch, TensorFlow, and Scikit-learn. Open to full-time roles in Germany.

View ExperienceView ProjectsContact
PythonSQLDeep LearningPower BIAWS
Kishan Sheladiya
2+ yrs
Experience
ML/AI
Focus
M.Sc.
Current

About Me

I am a Data Science and ML Engineer completing my M.Sc. in Mathematical Modelling, Simulation & Optimization at Universität Koblenz, Germany. I have hands-on experience building ML pipelines, ETL systems, and analytical dashboards across internships and personal projects.

My thesis focuses on GPU-accelerated time series classification using Optimal Transport (PyTorch/CUDA), benchmarked across medical, energy, and traffic datasets. I enjoy combining mathematical rigour with practical ML engineering to solve real-world problems. I am currently open to full-time roles in Data Science, ML Engineering, and Data Analytics in Germany.

GitHubLinkedInDownload CV
Primary Focus
Machine Learning & Data Science
University
Universität Koblenz
Current Degree
M.Sc. Mathematical Modelling, Simulation and Optimization
Thesis
Time Series Classification via Optimal Transport (PyTorch/CUDA)
Available From
October 2026
Location
Koblenz, Germany

Education

Academic background focused on mathematical modelling, ML/AI, and data-driven methods.

Master of Mathematical Modelling, Simulation and Optimization

Universität Koblenz • Koblenz, Germany

Oct 2022 – Present
Machine Learning and Data Mining, Artificial Intelligence, Optimization, Data Science, Big Data.
Details Image
Current

B.Tech. Mechanical Engineering

Vellore Institute of Technology • Vellore, India

Jul 2018 – May 2022
Python, Advanced Engineering Mathematics, Statistics, Operations Research, Data Analytics.
Details Image
Completed

Experience

Internships focused on ML engineering, analytics, automation, and turning data into production-ready insights.

Intern Machine Learning Developer

Digicode Informatik LLP • Gujarat, India

Jan 2022 – Jun 2022
  • Developed and maintained ML pipeline modules in Python (PyTorch, TensorFlow), contributing to data preprocessing, model training, and testing workflows.
  • Integrated REST API interfaces and AWS S3 storage into ML pipelines, supporting automated data retrieval and storage.
  • Gained exposure to code review processes and CI/CD workflows. Supported documentation of experiments and test results.
Company Logo
PythonPyTorchTensorFlowAWS S3REST APICI/CD

Data Analyst Intern

Kautilyum IT Services • Gujarat, India

Apr 2021 – Sep 2021
  • Built and maintained ETL pipelines and SQL queries to process data from multiple sources across client projects.
  • Created Power BI dashboards (DAX, Power Query) and conducted EDA and basic statistical analysis.
  • Ensured data quality and accuracy. Produced KPI reports to support the team.
Company Logo
PythonSQLPower BIDAXETLEDA

Master Thesis

Recently completed research project combining Optimal Transport, GPU acceleration, and real-world time-series classification.

May 2025 – Oct 2025

Time Series Classification by Optimal Transport Method

Universität Koblenz

Completed
  • Built a GPU-accelerated Optimal Transport Warping (OTW) model in Python (PyTorch) with CUDA for efficient time-series classification.
  • Benchmarked OTW vs. Dynamic Time Warping (DTW) on 7 UCR datasets, achieving higher accuracy in structurally complex domains.
  • Applied machine learning, optimization, and statistical techniques to real-world datasets across medical, energy, and traffic domains.
  • Developed a reproducible pipeline for experiments and benchmarking, connecting advanced mathematical methods with real-world applications.
PythonPyTorchCUDAOptimal TransportTime SeriesBenchmarkingClassification

Key Results

Datasets
7 UCR datasets
Domains
Medical, Energy, Traffic
Method
OTW vs DTW
Metrics are reported from thesis benchmarking (OTW vs. DTW).
gitHubView GitHub

Projects

gitHubView GitHub

Predictive Maintenance – Anomaly Detection in Sensor Data

Machine Learning • Personal Project • Oct 2024 – Jan 2025

  • Analyzed industrial sensor time series data to detect machine failures. Performed EDA and feature engineering on raw sensor inputs.
  • Trained and evaluated classification models (Random Forest, XGBoost) and an Autoencoder (PyTorch) for anomaly detection in imbalanced datasets.
PythonPyTorchXGBoostScikit-learnAnomaly DetectionTime Series
GitHub

Sales Forecasting – Time Series Analysis & Forecasting

Machine Learning • Personal Project • Jun 2024 – Aug 2024

  • Prepared and analyzed sales time series data using Python (Pandas, NumPy). Identified seasonality and trends via statistical decomposition.
  • Developed and compared forecasting models (ARIMA, XGBoost, LSTM) for future revenue prediction. Evaluated performance using MAE and RMSE.
PythonLSTMXGBoostARIMATime SeriesPandasNumPy
GitHub

Customer Churn Prediction – Classification & Business Analytics

Machine Learning • Personal Project • Mar 2024 – May 2024

  • Built a churn prediction model on customer data using Python (Scikit-learn, Pandas). Handled class imbalance using SMOTE.
  • Compared classification models (Logistic Regression, Random Forest, XGBoost). Visualized results and derived actionable business recommendations.
PythonScikit-learnXGBoostSMOTEData VisualizationBusiness Analytics
GitHub

Book Recommendation System

Machine Learning • Personal Project • Jan 2025 – Mar 2025

  • Built a book recommendation system using collaborative filtering, content-based filtering, and RAG with FastAPI.
  • Applied data preprocessing, feature engineering, and model evaluation using ROC AUC, F1-score, Precision, and Recall.
PythonFastAPIRecommender SystemsCollaborative FilteringRAGEDA
GitHub

ETL Epidemiological Data Analysis

Data Engineering • Personal Project • Jul 2024 – Sep 2024

  • Built ETL pipelines to collect and preprocess epidemiological data, ensuring data accuracy and reliability.
  • Created an interactive Power BI dashboard to visualize key trends and patterns for analytical reporting.
ETLData CleaningPower BIEDATrend Analysis

Credit Card Fraud Detection

Machine Learning • Personal Project • Mar 2024 – Apr 2024

  • Developed a fraud detection system on an imbalanced dataset using Logistic Regression, Random Forest, XGBoost, and Neural Networks.
  • Evaluated models using precision, recall, F1-score, and ROC AUC. Visualized results using Power BI.
PythonXGBoostImbalanced LearningScikit-learnPower BIModel Tuning

Skills

Tools and technologies I use to build ML systems, data pipelines, and production-ready applications.

Programming Languages

PythonSQLJavaC++

Frameworks & Libraries

NumPyPandasSciPyScikit-learnPyTorchTensorFlowFastAPIOpenCVCUDAMySQL

Machine Learning & AI

Deep LearningNatural Language Processing (NLP)Neural NetworksFeature EngineeringData PreprocessingTime Series AnalysisAnomaly Detection

Data Visualization

Power BITableauMatplotlibSeabornPlotly

Analysis Methods

Statistical AnalysisStatistical ModelingHypothesis TestingRegression/ClassificationForecastingTime-Series AnalysisData Quality ManagementEDA

Cloud & DevOps

AWS (S3, EC2, Lambda)DockerGit/GitHubCI/CD PipelinesREST APIs

Tools & Utilities

ETLOptimizationJupyter NotebookVS CodeLinux

Certificates

4 total

Microsoft Certified: Power BI Data Analyst Associate

Microsoft / Udemy

2023
Power BIAnalyticsDashboards

Scalable Machine Learning with Apache Spark

Databricks

2023
Apache SparkScalable MLDatabricks

Apache Spark Programming with Databricks

Databricks

2023
Apache SparkDatabricks

Python Programming Bootcamp

Udemy

2021
PythonProgramming

Languages

English – Fluent

German – B2 (Learning)

Contact

Feel free to reach out for Full time, internships, research collaboration, or ML/AI projects.

Get in touch

The fastest way is email. You can also connect via LinkedIn or check my work on GitHub.

Location
Germany
Email
kishansheladiya07@gmail.com
Phone
+49 176 86664054
LinkedInGitHubDownload CV

Quick message