Work / Solaris

Solaris

A backend API that predicts solar power generation from NASA POWER and Open-Meteo data, and finds the best panel orientation.

RoleMachine learning
StatusBuilt
TypeSolar power generation prediction
Code ↗

How it fits together

STEP 01NASA POWER + Open-Meteo data
STEP 02XGBoost model
STEP 03Azimuth optimization
STEP 04Generation forecast

What I did

  • Built the prediction API in Flask.
  • Trained an XGBoost model on NASA POWER and Open-Meteo data.
  • Added azimuth optimization to find the best panel orientation.

Stack

  • Flask
  • XGBoost
  • Python
  • TypeScript
← PreviousSugam Next →AI Code Security Platform