Chemical AI Plant Identifier Assistant -- 2

I am setting up an end-to-end AI lab assistant that will ultimately help me eliminate unwanted plant species and isolate promising alkaloids for cell, tissue, and rodent studies. The very first milestone is to teach the system how to identify plant species strictly from their chemical fingerprints.

Here is what I need you to build now:

• A model or pipeline that ingests raw or pre-processed data from techniques such as LC-MS, GC-MS, or NMR and matches the signature to the correct species.
• Tight integration with reputable existing chemical databases so the model can cross-reference compositions in real time.
• A lightweight interface (CLI, small web app, or notebook) that lets me upload spectra or tabulated peak data and returns the species name alongside a confidence score.
• Clear documentation of the data schema, preprocessing steps, and every dependency so I can reproduce the workflow inside my own lab environment.

Accuracy and transparency are critical: please include the metrics you will track and the validation approach you will use so we can judge whether the identifier is trustworthy before moving on to the elimination and alkaloid-isolation phases.

If you have previous work in cheminformatics, spectroscopy analysis, or ML-driven taxonomy, link me to it and outline how you would tackle database integration and model training.