AI Geospatial MVP for Analysis
Budget: ₹12,500 – ₹37,500 INR
I’m putting together a lean proof-of-concept that mines Satellite imagery and GIS layers with AI so I can validate a larger platform idea. The goal at this stage is pure data analysis: ingest the two data sources, run machine-learning workflows that surface patterns or anomalies, then surface the results through a minimal interface or notebook I can demo to stakeholders.
A typical stack for me would be Python with GeoPandas, rasterio, TensorFlow or PyTorch, plus PostGIS or a lightweight spatial database—but I’m open if you have a smarter path. What matters is that the pipeline reliably pulls in the imagery and vector layers, cleans and aligns them, applies the models, and outputs clear, reproducible insights.
Please attach past work that proves you’ve built something similar—ideally an AI application that touched either Satellite imagery, GIS data, or both. Code samples, short videos, or live links are perfect; lengthy proposals aren’t necessary at this point.
Deliverables
1. Working MVP repo (well-commented code, requirements file, and a concise README)
2. A simple UI, notebook, or dashboard that demonstrates the analysis end-to-end
3. Deployment/development guide so I can spin it up locally or on a cloud instance
I’ll review submissions based mainly on relevance of prior projects and how quickly you think we can hit a functional demo.
A typical stack for me would be Python with GeoPandas, rasterio, TensorFlow or PyTorch, plus PostGIS or a lightweight spatial database—but I’m open if you have a smarter path. What matters is that the pipeline reliably pulls in the imagery and vector layers, cleans and aligns them, applies the models, and outputs clear, reproducible insights.
Please attach past work that proves you’ve built something similar—ideally an AI application that touched either Satellite imagery, GIS data, or both. Code samples, short videos, or live links are perfect; lengthy proposals aren’t necessary at this point.
Deliverables
1. Working MVP repo (well-commented code, requirements file, and a concise README)
2. A simple UI, notebook, or dashboard that demonstrates the analysis end-to-end
3. Deployment/development guide so I can spin it up locally or on a cloud instance
I’ll review submissions based mainly on relevance of prior projects and how quickly you think we can hit a functional demo.