Hi, I’m Sudipto Chakraborty

Master’s Student Aerospace Informatics (University of Würzburg) Computer Science Graduate | Machine Learning | Software Engineering Würzburg, Germany

I build data-driven and software-intensive systems at the intersection of:

• Machine Learning & Data Science
• AI for Real-World Systems
• Remote Sensing & GeoAI
• Software Engineering (Python, C/C++)

I’m particularly interested in taking ML models beyond theory, building systems that are robust, interpretable, and usable in real environments.

• Foundation Models (e.g., Prithvi EO, Vision Transformers)
• Applied ML pipelines (end-to-end systems)
• Remote sensing + geospatial AI
• Efficient and clean system design

• Python, C/C++, JavaScript
• Git & GitHub
• Linux, Jupyter Notebook

• Regression, Classification, Clustering
• PCA, GMM, Feature Engineering
• Time-Series Analysis
• Model Evaluation & Validation

Libraries: Pandas, NumPy, scikit-learn, PyTorch, Matplotlib

• Remote Sensing & GeoAI
• Predictive Maintenance
• Computer Vision (Foundations)
• Aerospace Data Systems

Repository: Earth-Observation-using-Prithvi-EO-2.0

• Built a full pipeline combining:

  • Prithvi EO 2.0 embeddings (foundation model)
  • Spectral indices (NDVI, NDWI, SAVI, NDRE)
  • PCA + Gaussian Mixture Models (GMM)

• Performed unsupervised crop zone detection from satellite imagery

• Developed a multi-panel visualization dashboard for explainability

• Exported results as GeoTIFF for GIS workflows

Developed at Greenspin GmbH (Würzburg) → Data, imagery, and infrastructure provided → Full pipeline design and implementation done independently

Repository: aircraft-engine-predictive-maintenance

• Developed a machine learning pipeline to predict the Remaining Useful Life (RUL) of turbofan aircraft engines using the NASA CMAPSS multivariate sensor dataset.
• Built a leak-free predictive maintenance workflow with engine-wise train/validation splitting, missing value imputation, rolling-window features, and strict prevention of data leakage.
• Implemented and evaluated Dummy Regressor, Linear Regression, and Random Forest models, with Random Forest achieving the best performance (~18 RMSE).
• Designed the pipeline following industry-grade validation practices to ensure reliable time-series predictions for predictive maintenance applications.
• Generated Predicted vs. True RUL visualizations to analyze model performance, degradation trends, and prediction reliability for aerospace maintenance scenarios.

Repository: Autonomous-Camera-Trap-for-Insect-Monitoring

• Developed a fully autonomous edge AI insect monitoring system on Raspberry Pi 3 Model B+ for real-time insect detection and classification.
• Integrated a custom-trained YOLOv5n model capable of recognizing 33 insect classes from the IP102 dataset, performing all inference offline without cloud connectivity.
• Built a modular Python pipeline for image acquisition, preprocessing, AI inference, annotation, and CSV-based detection logging.
• Implemented a Flask-based web interface for real-time MJPEG video streaming and remote monitoring of the system.
• Designed for low-cost, field deployment with a custom 3D-printable enclosure, enabling long-term ecological and biodiversity monitoring.

• Developed a deep learning-based sign language recognition system to identify and translate hand gestures into readable text, improving communication accessibility.
• Built the application using Python, TensorFlow, Django, and MySQL, integrating computer vision and machine learning for gesture recognition.
• Implemented an end-to-end pipeline for image/video capture, preprocessing, feature extraction, and gesture classification using deep learning techniques.
• Leveraged AWS cloud services for scalable application deployment and efficient data management.
• Published the project as a research paper in the International Research Journal of Engineering and Technology (IRJET), May 2024.

• Earth Observation Models - Internship (Greenspin GmbH)
• AWS Cloud Internship – cloud fundamentals & deployment
• Full-Stack Development Training – React, Node.js
• ESRI ArcGIS Imagery MOOC – remote sensing & GeoAI

I’m currently looking for:

• Working Student / HiWi roles in:

  • Remote Sensing / GeoAI
  • Machine Learning / Data Science
  • Software Engineering
  • Applied AI Systems

I’m especially interested in hands-on roles where I can:

• Build real systems
• Work with experienced engineers/researchers
• Contribute to meaningful projects

• LinkedIn: https://www.linkedin.com/in/sudipto-chakraborty-4a96761b7/
• Email: sudiptopg@gmail.com
• Instagram: @sudiptochakraborty_

Thanks for visiting my profile!