Planex is an AI-powered exoplanet detection platform built during a 30-hour hackathon as part of the NASA Space Apps Challenge 2025.
It leverages NASA's open-source Kepler, K2, and TESS datasets to automatically identify potential exoplanets using machine learning.
- 🪐 AI/ML Model: Trained on NASA's Kepler, K2, and TESS datasets to classify exoplanet candidates.
- 💡 Automatic Classification: Identifies data points as Confirmed Exoplanets, Candidates, or False Positives.
- 🌐 Interactive Interface: Users can upload new datasets and view predictions in real-time.
- ⚡ FastAPI Backend: Efficient inference and preprocessing for large-scale astronomical data.
- 🛰️ Built in 30 Hours: Developed end-to-end during a 30-hour hackathon event.
- Dataset Input: NASA's open-source Kepler, K2, and TESS mission data are used for training and testing.
- Preprocessing: The backend cleans, scales, and prepares transit and orbital data.
- Model Inference: The trained model predicts whether each observation corresponds to a real exoplanet.
- Visualization: The frontend displays prediction confidence levels and categorized results.
Frontend:
Backend:
👉 Backend source code: github.com/Fr33b13101/planex