Menu

Hello, I'm Jayesh J. Pandey

AI/ML Engineer
Space Technology
Enthusiast

About Me

Computer Engineering student and NASA Space Apps Challenge 2025 Global Winner specializing in AI/ML, space technology, and geospatial intelligence. Proficient in PyTorch, full-stack development, and deploying scalable deep learning solutions for remote sensing, medical imaging, and computer vision.

Published researcher in generative AI and space weather prediction with active open-source contributions. Currently pursuing B.E. Computer Engineering at PVGCOE, Nashik (GPA: 8.70) with graduation expected in 2026. President of PVGCOE Nashik, hosting events, workshops, and guest sessions.

My expertise spans PyTorch, TensorFlow, React Native, FastAPI, and geospatial technologies like GDAL and Sentinel-2 processing. I've developed impactful projects including FaceParser (published on PyPI), SolarSim for CME detection, and marine debris detection systems achieving 78% precision.

My practical skills include AI/ML, space technology, and app development, with experience as a Machine Learning Engineer intern at CIT Internship Studio. I continue to explore emerging technologies, including Google’s Gemini AI, and am authoring a book titled "Life is All About 7 Minutes," inspired by spiritual philosophies.
Driven by curiosity and innovation, I aim to make a meaningful impact in software engineering and space technology while shaping the future of education and technology.

Work & Education

FEB 2026 - PRESENT

ISRO

Machine Learning Engineer

Develop PyTorch neural networks for Semi-Gaussian signal parameter identification using low-sampling data, optimize architectures through comparative analysis, integrate IoT pipelines, and advance AI-driven space technology research for real-time signal classification.

AUG 2025 - JAN 2026

EaseMeMed

LLMOps Engineer Intern

Built PyTorch pipelines for object detection using transfer learning and augmentation, implemented reproducible ML workflows, mastered LLM architectures, and optimized inference pipelines, reducing latency and resource usage effectively.

JAN 2025 - MAR 2025

Zensar Technologies

AI Engineer Trainee

Completed structured ML tasks on regression, classification, model tuning, and evaluation using metrics such as F1, ROC-AUC, and confusion matrices. Practiced feature engineering and model tuning on real-world style datasets with focus on explainability and business use.

SEP 2024 - OCT 2024

Internship Studio

Machine Learning Engineer Intern

Built multi-class classification models in PyTorch for visual multi-object scenarios, profiling bottlenecks to reduce training time by about 20%. Implemented experiments comparing architectures, optimizers, and learning-rate schedules, logging metrics for objective comparison.

AUG 2024 - NOV 2025

GSSOC

Open Source AI Contributor

Contributing to NLP and Computer Vision modules across multiple open-source repositories, focusing on model performance and modular, maintainable code. Implementing data preprocessing, augmentation, and evaluation pipelines that make projects easier for new contributors to extend. Using Git, issues, and pull requests to collaborate asynchronously with maintainers and other contributors.

2023 - 2025

Hackathon Achievements

NASA Space Apps Challenge 2025 Global Winner | SIH 2025 Winner | PVG HackerHub 2025 Winner

NASA Space Apps Challenge 2025: Global Winner for innovative space technology solution.
SIH 2025: Winner of Internal Hackathon with cutting-edge AI/ML solution.
PVG HackerHub 2025: Winner with innovative technical implementation.
GDG Hacktoberfest 2025: 2nd Runner Up for open-source contributions.
SIH 2023: Finalist for investigating vulnerabilities in crypto libraries for OpenVPN.

Aug 2022 - June 2026

PVGCOE & SSDIOM, Nashik

B.E. Computer Engineering — GPA: 8.70

Specializing in AI/ML, space technology, and full-stack development. Key coursework: Data Structures & Algorithms, Operating Systems, DBMS, Computer Networks, Artificial Intelligence, Machine Learning. President of PVGCOE Nashik student body.

July 2020 - June 2022

Satish Pradhan Dnyanasadhana College

Higher Education (HSC)

At Satish Pradhan Dnyanasadhana College, I pursued my higher education in 11th and 12th grades with dedication and excellence. Immersed in a dynamic learning environment, I excelled academically, securing an A grade. The college provided a supportive atmosphere and comprehensive curriculum, fostering my intellectual growth and preparing me for future challenges.

July 2007 - June 2019

HOFEHS

School (L.Kg to SSC)

HOFEHS, a renowned educational institution in Mumbai, offers a comprehensive journey from L.KG to 10th grade. With a commitment to academic excellence and holistic development, it provides a supportive environment where students thrive intellectually and socially. Experienced faculty, modern facilities, and a rich curriculum ensure a well-rounded education for every student.

Capabilities

NASA Space Apps Challenge 2025 Global Winner | AI/ML Engineering | Space Technology | Geospatial Intelligence

AI/ML & GenAI Engineering

Expert in PyTorch, TensorFlow, LLMs (Llama 3.2, RAG, Quantization via bitsandbytes/GGUF), and deep learning architectures including CNN, LSTM, Transformers, and GANs. Proficient in vLLM, LangChain, LangGraph, and deploying scalable ML solutions for production.

Space Technology

NASA Space Apps Challenge 2025 Global Winner with expertise in space weather prediction, CME detection using Aditya-L1 and SOHO satellite data, and geospatial intelligence for space applications.

Geospatial Analysis

Proficient in remote sensing, multi-temporal change detection, Sentinel-2 & Landsat processing, glacial lake and road extraction using GDAL, GeoPandas, and advanced geospatial tooling.

Full-Stack Development

Expert in React Native, FastAPI, Express.js, Node.js with experience in building scalable applications. Published Python packages on PyPI and developed cross-platform mobile applications.

Research & Publications

Published researcher in generative AI and space weather prediction. Authored papers on architectural design using deep learning and CME detection using satellite data, with work under review in peer-reviewed venues.

Open Source Contribution

Active contributor to AI/ML repositories with focus on NLP and Computer Vision modules. Published FaceParser library on PyPI for facial feature parsing using BiSeNet architecture with GPU acceleration.

Selected Works

Here are some of my selected works I have done lately. Feel free to check them out.

FaceParser: Face Parsing Library

Python library for facial feature parsing using BiSeNet architecture, supporting segmentation of 19+ facial regions with GPU acceleration. Published on PyPI.

Python, PyTorch, OpenCV, CUDA, NumPy, BiSeNet Architecture

SolarSim: Advanced CME Detection

Deep learning system for detecting Coronal Mass Ejections (CMEs) using Aditya-L1 and SOHO satellite data, with desktop interface for space weather prediction.

PyTorch, CNN-LSTM Architecture, Aditya-L1 Data, SOHO Data, Tkinter GUI, Space Weather Prediction

Swap Health – AI Fitness & Health App

React Native application combining pose estimation and AI assistance to guide users through workouts with real-time feedback on 25+ exercise types.

React Native, FastAPI, MediaPipe, Hugging Face, REST APIs, Pose Estimation, AI Chatbot

Btechnotes

Web Design

High-impact educational platform targeting engineering students, reaching 1M+ views in the first year. Built and maintained end-to-end including deployment, content structuring, and performance optimization.

Graph-Based-Query-System

AI Knowledge Graph

Full-stack intelligent dashboard for Order-to-Cash data. Automatically builds a Neo4j knowledge graph from JSONL documents and enables natural language queries via a safe LangGraph + LLM agent.

Automated Geospatial Feature Detection

Geospatial AI, Remote Sensing

Developed semantic segmentation and change detection models using GDAL on Sentinel-2/Landsat imagery; built React Native mobile app for real-time visualization of glacial lakes and roads.

Marine Debris Detection (PlanetScope)

Deep learning model for detecting marine debris from PlanetScope satellite imagery achieving 78% precision and 70% recall for ocean cleanup monitoring.

TensorFlow, PlanetScope Imagery, Object Detection, Satellite Image Processing, Environmental Monitoring

Ekip Bhaskar – CME Early Warning

Space Weather AI

AI-driven early warning system for Halo CMEs using ISRO's Aditya-L1 SWIS data. Achieves 84% accuracy, validated against CACTUS database, with kinematic modeling and React Native mobile visualization.

Lunar Crater Detection

Satellite Image Processing

End-to-end ML pipeline for lunar surface feature detection using LRO/Chandrayaan datasets. Covers data prep, CNN+FPN model training, experiment tracking, and production-ready model artifacts.

TrainIQ — Universal ML/DL Library

End-to-end ML/DL pipelines with a single call. Auto model selection, hyperparameter tuning via Optuna, and deployment scaffolding for tabular, image, text, and time-series data.

PyTorch, TensorFlow, XGBoost, BERT, ResNet, Optuna, ONNX, FastAPI, Ollama, HuggingFace LLMs

Get In Touch

jayeshpandey754@gmail.com

+91-8087967911

NASA Space Apps Challenge 2025 Global Winner | AI/ML Engineer specializing in space technology and geospatial intelligence. Let's collaborate on innovative AI solutions for space, healthcare, and environmental applications. Email Me.