class GeospatialScientist:
name = "Athithiyan M R"
role = "Geospatial Data Scientist"
based_in = "Tamil Nadu, India"
org = "Auroville Consulting"
skills = ["Satellite Remote Sensing", "Deep Learning (U-Net, CNN)",
"Land-Use Classification", "Climate Analytics", "GIS"]
stack = ["Python", "Google Earth Engine", "QGIS", "PyTorch",
"FastAPI", "Docker", "PostgreSQL/PostGIS"]
passion = "Turning satellite imagery into climate-actionable intelligence"Geospatial data scientist with 4+ years of professional experience building end-to-end pipelines — from raw Sentinel-2 imagery to policy-ready maps that feed into state-level planning decisions. My work spans 15,000+ satellite scenes, 40+ geospatial layers, and deep learning models deployed for land-use classification, plantation mapping, and urban climate analysis across Tamil Nadu.
Currently at Auroville Consulting, collaborating with the Tamil Nadu State Planning Commission on blue-green infrastructure planning and climate resilience.
Seeking fully remote roles in geospatial AI, climate analytics, or earth observation at research organizations, climate-tech companies, or think tanks.
- 🌿 Blue-Green Infrastructure Framework — State-wide ecosystem services prioritization for Tamil Nadu (with State Planning Commission)
- 🌴 Coconut Plantation Mapping — U-Net deep learning pipeline on Sentinel-2 for Coimbatore, with a collaborative verification tool
- 🏙️ Chennai Urban Climate — Built-up area extraction using deep learning for heat island and climate vulnerability analysis
- 📡 Exploring Google & Tessera embeddings for improved satellite image segmentation
|
Auroville Consulting × TN State Planning Commission Regional-scale geospatial analysis to map and prioritize Blue-Green Networks for flood regulation, carbon storage, and biodiversity.
|
Deep learning pipeline (U-Net) for coconut plantation mapping from Sentinel-2 imagery using Descals et al. (2023) training labels.
|
|
Deep learning pipeline (U-Net) for built-up area extraction from Sentinel-2 imagery — applied to Chennai for urban heat island analysis.
|
Auroville Consulting · Aug 2022 – Oct 2023 Research and spatial analytics supporting Tamil Nadu's net-zero energy transition.
|
|
Multi-use suitability mapping of unused land for forestation, agriculture, water harvesting, housing, industry, and solar.
|
Sentinel-2 Built-Up Pipeline End-to-end ML pipeline for built-up area identification Lila CRS Converter Full-stack geospatial web app (FastAPI + Streamlit), Dockerized with CI/CD Coconut Verification Tool Collaborative web tool for verifying coconut training labels across 38 districts Bill Calculator 25-year economic analysis tool for Solar + Storage systems |
| Domain | Technologies |
|---|---|
| 🛰️ GIS & Remote Sensing | QGIS · ArcGIS · Google Earth Engine · Sentinel-2 · Landsat |
| 🤖 Machine Learning | PyTorch · TensorFlow · Scikit-Learn · U-Net · CNN · Random Forest · XGBoost |
| 🐍 Geospatial Python | GeoPandas · Rasterio · GDAL · NumPy · Pandas · Matplotlib |
| 🗄️ Database | PostgreSQL · PostGIS |
| 🌐 Web & Backend | FastAPI · Streamlit · REST APIs · JavaScript |
| ⚙️ DevOps | Docker · GitHub Actions · Linux · Bash · Self-hosted VPS |
2026 🌿 Blue-Green Infrastructure Framework — TN State Planning Commission
│ 🌴 Coconut Plantation Mapping — U-Net on Sentinel-2
│ 🏙️ Chennai Urban Climate — Deep Learning for Built-Up Extraction
│
2025 🗺️ Kallakurichi Land Suitability Assessment — Multi-use mapping
│
2024 ☀️ SET-TN — Sustainable Energy Transformation, Tamil Nadu
│ Land suitability for solar, tariff modelling, policy analytics
│
2022 📊 Joined Auroville Consulting as Geospatial Data Analyst
│
2022 🎓 Master in Data Science — GUVI (IIT Madras)
Open to collaborations in geospatial AI, climate analytics, and open-source environmental tools.