Skip to content

Modular ontologies

Your business has a language. Percidian makes it usable by AI.

A shared business foundation, the domain ontology the work needs — Legal, M&A, Investment, Finance, or Biotech — and your organization’s own vocabulary, in one model every approved AI system draws on, so a filing, a finding, or a safety signal means what it means inside the work.

Available domain ontologies

Five available domains, each an explicit chain of entities AI can follow.

Legal · 9 entities in the chain

  1. Client
  2. Matter
  3. Party
  4. Claim
  5. Filing
  6. Obligation
  7. Deadline
  8. Precedent
  9. Strategy

What AI can follow. A filing belongs to a matter, a deadline follows from an event, and prior firm work may be relevant — subject to matter and document access.

Generic intelligence, specific work

Generic AI knows the language. Percidian knows what it means inside the work.

A model recognizes the word “matter.” It does not inherently know how that matter connects to a client, responsible partner, opposing party, filing, precedent, deadline, obligation, confidentiality boundary, and prior strategy.

Percidian makes those concepts and relationships explicit, maintains them across the enterprise, and makes the relevant authorized view reusable across every approved AI system.

Composed, not built

An ontology is the foundation, the packs a practice needs, and your own vocabulary.

The ACE Foundation carries what every business shares — people, organizations, documents, communications, decisions, obligations, permissions. Data packs layer the substance of a domain on top of it, and an ontology composes from them: the Legal ontology takes the Legal, Legal tech, and Business packs; M&A takes M&A, Legal, Finance, and Business. Available today: Legal, M&A, Investment, Finance, and Biotech; the foundation is built to support additional validated domains.

The work does not stay inside one category. A transaction can involve corporate entities, agreements, revenue, customers, employees, products, intellectual property, regulatory obligations, diligence findings, and decisions. Every pack extends the same foundation and declares what it depends on, so AI can follow a relationship across domains without crossing into another knowledge base.

Legal pack — Matter, Party, Claim, Filing, Obligation, Deadline, Precedent · 9 types · depends on coreLegalLegal tech pack — Document system, workflow, e-discovery, billing entities · 6 types · depends on core, legal, techLegal techM&A pack — Target, Deal, Workstream, Diligence finding, Exposure, Condition · 9 types · depends on core, legal, fin, bizM&AFinance pack — Instrument, Holding, Exposure, Valuation, Covenant, Forecast · 16 types · depends on core, legalFinanceFintech pack — Account, Payment, Ledger, Rail, Provider · 6 types · depends on core, legal, finFintechBusiness pack — Company, Unit, Product, Customer, Contract · 5 types · depends on coreBusinessSales pack — Account, Opportunity, Quote, Renewal, Owner · 7 types · depends on core, biz, finSalesPeople & HR pack — Role, Employment, Compensation, Policy · 6 types · depends on core, finPeople & HRLife sciences pack — Program, Compound, Study, Site, Endpoint, Safety signal, Submission · 8 types · depends on core, legalLife sciencesTechnology pack — System, Service, Incident, Release, Dependency · 7 types · depends on core, legalTechnologyACEFoundation

Your vocabulary becomes infrastructure

Make the way your organization works consistent across every AI system.

Matter types, transaction stages, product hierarchies, risk categories, committee structures, client classifications, and internal definitions — part of the model, not recreated in every prompt or hard-coded inside every application.

Restated in every prompt

  • Definitions repeated per prompt, per user
  • Taxonomies hard-coded inside each application
  • Meaning drifts between tools and teams
  • Every new workflow starts from a blank vocabulary

Defined once in the model

  • One definition, in the organization layer
  • Shared by every approved assistant, agent, and application
  • Consistent, reusable meaning across the enterprise
  • New workflows reuse the meaning already defined

Teach every approved AI system how your business works.

Combine deep domain understanding with the language, structure, and operating reality unique to your organization.

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.

See it on your business.

A briefing mapped to your systems, your permissions, and one workflow that matters.

A Percidian solutions lead will follow up to schedule the briefing. By submitting this form, you agree to our Privacy Policy.