ChemAgent. It doesn’t guess. It calculates.
An example of our domain-specific model and agent development. It connects chemistry knowledge to scientific sources and calculation tools, and can be adapted to research and analysis tasks.
- The question: A research question, in plain words.
- The molecule: It writes α-pinene as SMILES, then folds it into its structure.
- The answer: Calculated, not guessed: 100 g of α-pinene burns 328.8 g of oxygen.
ChemAgent is asked how much oxygen 100 g of α-pinene uses as it burns. It writes the molecule as SMILES, CC1=CCC2CC1C2(C)C, folds it into its structure and calculates the answer: 328.8 g of oxygen.
Molecules, sources and calculations in one research workflow.
ChemAgent is ALTAI’s AI product for chemistry and scientific analysis. It connects a chemistry-focused model to scientific data sources and calculation tools. We adapt it to your team’s molecule review, literature research and data analysis tasks.
Work with molecular data
Resolve common chemical identifiers, retrieve PubChem and ChEMBL records, and calculate selected molecular properties with tools such as RDKit.
Search scientific sources
Search sources such as PubMed, OpenAlex and Europe PMC, then present the publications and available reference details for review.
Run calculations and data analysis
Use Python, Pandas and DuckDB-based tools for chemical calculations, queries over CSV and JSONL data, and statistical analysis.
Where it fits.
| Team | The task | What changes |
|---|---|---|
| Chemistry R&D | Collect a compound's identifiers, selected properties and scientific database records. | Prepare a preliminary review with sources for a researcher to check. |
| Scientific information teams | Search supported sources for publications relevant to a focused research question. | Reduce source gathering work and keep the publications behind a synthesis visible. |
| Quality and process analytics | Examine production data and run agreed statistical calculations. | Prepare a repeatable analysis and draft report for expert assessment. |
Set up around your work.
ALTAI connects your datasets and develops the scientific tools, calculation steps and reporting interface the task needs. We scope the work around a research or analysis task and evaluate results against examples reviewed by your team. Outputs receive expert review before scientific decisions; any necessary laboratory validation is performed separately.
We measure together
- Calculation and unit correctness on reviewed examples
- Source relevance and search coverage
- Expert correction time per analysis
Spec sheet
- Category: Domain-specific AI product for chemistry
- Role: ChemAgent is a product of our domain-specific model and agent development work. Chemistry knowledge, scientific sources and calculation tools work together in one research workflow; results require expert review and laboratory validation where needed.
- Model: An expert model trained for chemistry
- Data: CSV, JSONL, your organisation’s sources
- Review: Experts check results; the lab confirms where needed.
- Works with: Yada, isanagent