Geospatial Machine Learning Scientist

Job Title: Geospatial Machine Learning Scientist

Location:
Bangalore, India

Experience:
4+ years in geospatial data science and ML, with expertise in multi-source dataset integration and high-accuracy environmental modeling.

About Alt Carbon:
Alt Carbon is India's leading Carbon Removal (CDR) company. We blend science, technology & tradition to undertake climate action to turn the clock on historic emissions in the atmosphere. Our science and technology takes place in Bangalore; our operations are undertaken in Darjeeling's heritage tea estates.

Alt Carbon was started by sibling tea planters Shrey and Sparsh Agarwal, in order to rescue their family's tea estate from being at-risk of climate impact to becoming the frontier of climate action.

Read about our journey below:

Our Journey to Revive Darjeeling

To read more about our work, check out this news feature about us, go through this twitter thread, or have a look at our work in Darjeeling.

Role Overview:

We are looking for a Geospatial Machine Learning Scientist to do the research and development of models that integrate multi-source geospatial datasets with our proprietary soil dataset. Your work will directly contribute to building high-accuracy environmental insights for climate and agricultural applications.

You will be responsible for researching, curating, and preprocessing publicly available datasets (e.g., multispectral satellite imagery, atmospheric data, land cover maps, soil and vegetation indices), integrating them with proprietary datasets, and designing ensemble ML pipelines that extract actionable information.

What You'll Do:

Data Discovery & Integration

• Identify and evaluate publicly available geospatial datasets (e.g., NASA, ESA, USGS, ISRO, FAO, Copernicus).
• Integrate multispectral, hyperspectral, atmospheric, and soil datasets into a unified data model.
• Align datasets with varying spatial, temporal, and spectral resolutions.

Model Development

• Design and train ensemble ML models leveraging multiple data modalities (satellite imagery, soil chemistry, weather patterns).
• Apply deep learning (CNNs, Vision Transformers) and classical ML techniques for classification, regression, and prediction tasks.
• Experiment with fusion architectures that combine multispectral imagery with tabular soil and climate data.

Scientific Research & Validation

• Benchmark models against state-of-the-art methods from peer-reviewed literature.
• Collaborate with soil scientists, climate experts, and agronomists to validate model outputs.
• Document methodologies and publish technical reports or research papers when applicable.

Deployment & Scaling

• Optimize models for large-scale inference on cloud infrastructure.
• Build reproducible data pipelines for ongoing dataset updates and retraining.

What We're Looking For:

• MSc/PhD in Geoinformatics, Remote Sensing, Machine Learning, Environmental Data Science, or related field.
• Strong experience with
  geospatial data
  (Sentinel, Landsat, MODIS, ERA5, SMAP, etc.) and formats (GeoTIFF, NetCDF, shapefiles).
• Proficiency in
  Python
  , with experience in ML/DL frameworks (PyTorch, scikit-learn, XGBoost, LightGBM).
• Familiarity with geospatial processing libraries (GDAL, rasterio, geopandas, xarray, rioxarray).
• Experience in
  multi-modal and ensemble ML models
  .
• Understanding of soil science, climate modeling, or agricultural data is a plus.

Preferred Qualifications:

• Experience with cloud geospatial platforms (Google Earth Engine, AWS Open Data).
• Publication record in remote sensing or ML journals/conferences.

Why Join Us?

There are innumerable JDs floated everyday in Bangalore. How many of them will offer you an opportunity to work on something as challenging, meaningful and important as Climate Change? And how many of them have operations in a place as beautiful, legendary and special as Darjeeling?