Moloracle Engine Early Access
Early access is intended to begin with a concrete workflow, not a generic demo. We want to understand the input data, the decision a team is trying to improve, and the evidence required to judge whether the workflow helps.
What the pilot includes
A pilot can cover dataset preparation, model inference, candidate review, and export of the artefacts needed for a follow-up calculation or experiment. The exact scope depends on the workflow and the hardware boundary the team needs to preserve.
Each pilot should have a defined success criterion and a review checkpoint. Faster inference is useful only when it reduces repeated work without hiding important assumptions.
A sensible pilot starts with a bounded dataset and an agreed input contract: which structures are included, how salts and tautomers are handled, whether stereochemistry is trusted, and which experimental conditions are available. That makes a later comparison meaningful instead of turning the pilot into an open-ended demo.
What early access is not
It is not a universal accuracy claim, a replacement for experimental validation, or a promise that every model is production-ready. Private alpha is the stage for learning where the workflow is reliable and where it still needs refinement.
For chemistry and biology, validation remains tied to the property being measured. A ranking can be useful even when the absolute value is uncertain, but the decision rule should say what happens next: a repeat calculation, a control compound, a binding assay, a polarimetry measurement, or a different model.
Who should apply
Research and development teams with a defined molecular decision, an understanding of their data boundary, and the capacity to review model outputs alongside existing chemistry or biology practice are the best fit for the first cohort.
The strongest starting point is therefore not the largest dataset. It is a narrow workflow with identifiable inputs, a measurable review burden, and enough domain expertise to distinguish a useful prioritisation from a chemically implausible suggestion.