For biotech
Give AI the full scientific and business context behind every program.
Percidian connects programs, assets, targets, indications, studies, sites, endpoints, evidence, safety signals, submissions, milestones, partners, decisions, and dependencies into one governed, permissioned understanding — so every approved AI system sees what changed, what it affects, who owns the next decision, and which evidence supports it. A finding on a shared mechanism, vendor, site, or supplier lands on every program that depends on it.
Program and partner boundaries applied before context reaches AI · Every point cited to its source
Safety signal · ALT elevation · class effect
- seen in
- BX-214 · Phase 2 safety review
- shares
- Mechanism with BX-301 and BX-118
- reaches
- 3 of 4 programs
- moves
- Two milestones, one partner report
- owner
- Safety review board · this month
- evidence
- Review minutes · protocol · site data
Illustrative exampleOne signal, resolved to every program it touches.
Shared across programs
One safety signal. Every program it touches.
A single-asset company reads a safety review as one program’s problem. A portfolio cannot afford to. One shared dependency, every program it reaches.
Safety signal · ALT elevation · class effect. Seen first in BX-214’s Phase 2 review. BX-301 and BX-118 share the mechanism, so the safety review reaches three programs — not one. Reaches 3 of 4 programs.
Four programs — BX-214 (Phase 2), BX-301 (Phase 1), BX-118 (Phase 3), BX-407 (Preclinical) — and four shared dependencies: Safety signal, ALT elevation · class effect, touching BX-214 and BX-301 and BX-118; Vendor, Meridian Clinical · CRO, touching BX-214 and BX-118; Site, Site 14 · enrolling, touching BX-301 and BX-118; Supplier, Drug substance · single source, touching BX-214 and BX-407. Now lit: Safety signal · ALT elevation · class effect. Seen first in BX-214’s Phase 2 review. BX-301 and BX-118 share the mechanism, so the safety review reaches three programs — not one.
Illustrative example
The brief
Move from new evidence to the next decision.
“What could put BX-214’s next development milestone at risk, and what decisions are required this month?”
- Enrollment timing and a supply dependency place the milestone date under pressure.Recruitment trend · vendor update · protocol timeline
- The delay affects the regulatory interaction, budget phasing, and the partner reporting obligation.Regulatory correspondence · budget · partner agreement
- Confirm the revised operating plan, approve the mitigation owner, and align the external milestone communication.Prior decision record · program plan
- Open: whether the endpoint window remains achievable under the proposed protocol change.Endpoint definitions · protocol amendment draft
Assembled from the latest study results, safety review, site performance, protocol amendments, regulatory correspondence, supply dependency, partner obligations, budget assumptions, and the prior decision record.
Illustrative example
One program memory across AI
Keep the program intact across every approved AI experience.
Research agents, clinical workflows, regulatory assistants, diligence tools, and executive AI draw from the same governed program understanding. Teams do not reintroduce the asset, rebuild its history, or reconcile competing versions of the evidence every time they change applications.
Approved source material becomes persistent program understanding that works across AI applications without repeatedly loading complete histories — inside each team’s permissions, with every point linked to its source.
- Persistent program memory
- Cross-functional context
- Role-specific permissions
- Source-linked evidence
The biotech ontology
Built on the available Biotech ontology, extended by your science and operating model.
Your naming, development stages, evidence standards, and governance extend it.
The available Biotech ontology models programs, assets, compounds, targets, mechanisms, indications, studies, protocols, cohorts, sites, investigators, endpoints, results, evidence, safety signals, submissions, regulators, milestones, vendors, partners, intellectual property, obligations, decisions, owners, and source records. The organization’s own naming, development stages, evidence standards, governance, portfolio model, and decision process extend that shared structure without creating another silo.
Bring your business into every AI decision
Give every approved AI system the context to do its best work.
Connect the knowledge your organization already has. Preserve what it learns. Apply the permissions it already trusts. Make that understanding available wherever people and agents work.