Welcome to the Spectochem organization! We are a research-driven team dedicated to bridging the gap between Machine Learning and Computational Chemistry.

🔬 Our Mission

The traditional pipeline for obtaining molecular absorption spectra relies on expensive and time-consuming quantum chemical calculations, such as Density Functional Theory (DFT). Our goal is to develop highly efficient, ML-based "virtual filters" that can rapidly estimate spectrograms and predict the physical-chemical properties of molecules without the massive computational overhead.

🧪 What We Do

• Accelerating Discovery: We train advanced predictive models to screen tens of thousands of chemical compounds, specifically focusing on identifying potential photocatalysts.
• Graph & Molecular ML: We leverage spatial geometries and graph representations using cutting-edge molecular machine learning frameworks.
• Open Science: We believe in trustworthy, transparent, and reproducible AI. All our final models, utility scripts, and pipelines are released under permissive open-source licenses (MIT/BSD).

---

👥 Contributors: Tymoteusz Zapała, Katsiaryna Viarenich, Radosław Woźniak, Nikodem Świerkowski.
🤝 Collaborating with Faculty of Chemistry experts to solve real-world chemical computing challenges.