The same workflow registry that produces a regulatory brief produces a discovery dossier and a patient-context therapeutic recommendation. The intent object changes; the engine does not. This is the platform under the practice you see in the engagements pages.
A regulatory question, a discovery question, and a patient-context question all enter the same engine through the same intent abstraction. They flow through the same workflow registry, pick up the same review discipline, and exit as the deliverable shape that fits.
An intent object describes the question and the context it lives in. The engine reads the same shape across regulatory, discovery, and patient-context engagements.
A pathway, a precedent, or a submission-strategy question scoped to an asset, an indication, and a jurisdiction. Reads against the regulatory record across FDA, EMA, PMDA, MHRA, NMPA.
A target, a hit, a scaffold, or an asset to evaluate — with the science context and the downstream-development boundary. Reads against the target, chemistry, and IP record.
A patient's specific mutation profile, prior-line history, clinical context, and the regulatory pathway available to act on the answer — from a molecular tumor board or specialty-clinic precision-medicine program.
The intent object is the input shape. Across all three lanes, it is the same abstraction. New lanes — population health, payer / HTA evaluation, regulatory-portfolio strategy — extend the schema without rebuilding the engine.
The engine is a registry of versioned analytical workflows. Each workflow is auditable end-to-end — its rev is logged on every claim it produces, and a reviewer can replay any output against the cited primary sources.
The intent object is matched to the workflows the question requires. A regulatory pathway question runs the regulatory-pathway workflow against the cited source set; a target validation question runs the validation workflow; a patient-context question runs the re-rank workflow against the patient-specific intent and escalates to analog or novel design where needed. Same engine, different selection.
The platform's capability is not the workflow library. It is the abstraction that lets the same engine pick up a new workflow — from a CRL pattern reading to a deal precedent reading to a patient-context re-rank — and run it against any intent type, with the same review discipline and audit trail. New workflows extend the registry without rebuilding the practice.
For any intent that asks "what therapeutic fits this target or this patient," the platform answers across an escalation ladder. This is a platform capability, not a service-tier description.
The entire known therapeutic space — approved drugs, clinical-stage compounds, literature analogs — scored against the target conformation and intent context. Default entry point for any patient or program intent.
When the known space falls short, propose minimal-modification analogs of the closest scaffold — predicted profile, synthesis route, IP positioning. Engagement escalates from Tier 1 only on demand.
When no precedent fits, generate novel matter against the target — virtual screen, generative chemistry, predicted ADMET and synthesizability. Hand-off package ready for experimental validation.
Genomic diagnostics tell you what is wrong at the patient level. The molecular design layer tells you what to do about it at the molecular level. One sells data. The other sells decisions — ranked, escalating, defensible.
The engagements visible today are the practice's first surface. The platform underneath was built so the same engine can absorb new intent types and new workflow domains without rebuild. A trajectory, not a roadmap.
Intelligence Studio, Discovery Engine, and patient-context engagements via molecular tumor boards and specialty clinics — all running the same workflow registry.
Patient-context re-rank at production rev. IP & FTO at rev 2. Failure-record and trial-design precedent at rev 1. The engine reads more questions, faster.
Population-health intent (cohort-context re-rank), payer / HTA intent (reimbursement-context evaluation), regulatory-portfolio intent (multi-asset pathway optimization). Same engine, new lanes.
The marginal cost of a new intent type or a new workflow is the cost of authoring the workflow, not rebuilding the platform. The practice grows by extending the registry.
Trajectory disclosure is honest. The engagements live today are real. The extensions named above are scoped, not shipped. The platform was built for the trajectory; the practice ships against it.