Use cases
Analyses stuck in today's backlog, right inside your AI assistant.
Example workflows from retrosynthesis proposals to stability fits, each running in Claude, ChatGPT or Cursor. Plus computed examples with input, tool and date.
Create an account, get your API key, use the tools in Claude, ChatGPT or Cursor.
Example workflows
From target molecule to salt form.
Retrosynthesis in chat
Claude analyses the target molecule, proposes routes, checks druglikeness and ADME profiles, and exports the list of proposed building blocks. Routed through ChemReason for retrosynthesis and CovaChem for ADME.
Plan an ICH stability study
Generate study plans based on ICH Q1A, fit Arrhenius models and derive shelf-life estimates, straight from a ChatGPT tab. Routed through the CovaStab suite.
Docking screen from Cursor
From a Cursor notebook: load a SMILES list, run a docking pipeline, rank hits by docking score and synthesisability. Routed through the CovaDock and CovaFold suites.
Formulation optimisation with DoE
Formulation optimisation via RSM designs in Copilot: proposal, data-collection plan, model fit, robustness check and recommendation, all through tool calls. Routed through the CovaDoE and CovaOpt suites.
Antibody stability profile
From a sequence: hydrophobicity patches, aggregation risk, pI, viscosity hotspots. Routed through the CovaBio and CovaStab suites.
Solubility screen for salt selection
pH-dependent solubility curves, antisolvent strategies, mixture predictions for salt form selection. Routed through the CovaSolve suite.
Proof
Measured, not claimed.
- AI models alone
- 14 to 41%
- With CovaSyn tools
- 76 to 92%
Applications
Computed examples from drug development.
Every page carries a real result computed with our own tools, with input, tool name and run date. Technically reviewed before it appears.
ICH M7 pre-screening of nitrosamine impurities: what the class actually tells you
Four computed ICH M7 classifications, including the case where the structural alert and the statistical model disagree. With TTC threshold, mechanism, and what the class means for your next decision.
Source-verified, 2026-08-03
See the calculationMolecular formula from exact mass: why 5 ppm is often not enough
Two computed examples show the difference between an unambiguous and an ambiguous assignment. With mass error in ppm, RDBE, and the question of which candidate formula you may discard.
Source-verified, 2026-08-04
See the calculationPredicting 1H shifts before you measure: what the confidence interval tells you
Nine computed 1H shifts for paracetamol, each with a confidence interval. Why the width of the interval matters more for planning than the predicted value.
Source-verified, 2026-08-04
See the calculationPlanning an ICH Q1A stability study: 33 determinations for a 24-month shelf life
A computed study plan for climatic zone II with conditions, timepoints and sample count. What the number 33 means for your lab planning and where it still changes.
Source-verified, 2026-08-03
See the calculationHPLC starting method for an acidic compound: which gradient, and why TFA
A computed method proposal for ibuprofen with column, eluents, gradient and predicted retention time. Which parts are a starting point, and which you will adjust at the instrument anyway.
Source-verified, 2026-08-03
See the calculation
By role
Looking for your role?
Head of Analytical / QC
Spectra and purity questions pile up, and sign-offs wait.
Learn moreCMC and formulation
Stability packages and shelf-life estimates take days, and DoE lives in Excel.
Learn moreQA and regulatory
ICH M7 assessments take weeks, and there is no quick first read on nitrosamine risk.
Learn more
FAQ
Frequently asked questions
Which use case is the best starting point?
For CDMOs and pharma CMC teams, ICH stability studies are the fastest entry: defined outputs and clear audit requirements. For medicinal chemistry and drug discovery: retrosynthesis or docking. For formulation teams: DoE. It makes sense to start with one use case and then expand.
How do I integrate CovaSyn into Claude, ChatGPT or Cursor?
Add the MCP server entry to your AI assistant's configuration, paste your API key from the CovaSyn dashboard, done, no code. Example configurations for Claude Desktop, Cursor, VS Code and Python are MIT-licensed on GitHub (covasyn-mcp-examples).
My use case is not listed. What now?
The examples come from discovery and CMC. Our tools cover discovery, analytics and CMC, from NMR and MS to stability and toxicology, so usually something fits. Describe your workflow in the form below; if you leave your email address, we come back with concrete tool suggestions.
Can confidential projects run in your own data centre?
Yes. If your IT security or compliance rules out external hosting, the MCP server runs as a container on your infrastructure, behind your firewall, and your data does not go to CovaSyn. We set this up as Done-for-You: Systems. We clarify the details in an initial call.
What does the audit trail for ICH M7 and ICH Q1 workflows look like?
Every call is recorded with its tool version, configuration and timestamp, without permanently storing your inputs and results. That keeps every result reproducible.
Tool missing?
Tool not listed, or could not find a fit?
Tell us briefly what you need. If we have it, we come back to you with the matching tool. If we do not, we check whether it fits into our MCP, and you are the first to hear.
Email is optional. By entering it, you consent to us replying to your request; the data is used only for that reply.
Next step
Try the first workflow on your own structures.
Create an account, get your API key, use the tools in Claude, ChatGPT or Cursor.