Reading composition from spectra
Could we cut the time it takes to interpret a spectrum?
We think reading a spectrum leans heavily on a trained eye. We would like to study a tool that shows peaks and likely components with its evidence, so interpretation takes less time. People make the final call, and we want anyone to be able to try the tool.
IR, Raman and UV-Vis spectra
We aim to connect and clean the spectra, then have agents read the peaks and verify each other.
Likely components and anomalies, with evidence
| Field | Chemistry |
|---|---|
| Status | Planned |
| Input | IR · Raman · UV-Vis spectra |
| Output | Component estimate · anomaly detection |
| Delivery | Joint research |