All projects

Climate

Solar Registry Screening

An explainable pipeline that turns solar installation records into a ranked shortlist.

The project

The pipeline validates registry exports, applies qualification filters, and ranks the remaining installations with explicit scoring rules. Each result includes the points and reasons behind its score.

My contribution

I built ingestion and validation, deterministic ranking, optional geospatial enrichment, and reporting. The public example uses five fictional installations and a fixed scoring date so its results can be reproduced.

Context

Python, Pydantic, PostgreSQL, and optional model-assisted scoring.

The public repository contains synthetic examples. The scores illustrate a screening method; they are not investment recommendations.

Next project: DSTL Chat