AI and automation architecture
GAIA Civil uses a coordinated AI workflow layer to process project records while preserving review, recovery, and audit boundaries.
Reference version 1.1 · Updated September 4, 2026
System responsibilities
ORION is the internal name for the coordinated AI workflow layer that powers GAIA Civil’s document analysis. It is a technical component of GAIA Civil, not a separate customer product.
- Receive an authorized, tenant-scoped project record.
- Create text, rendered-page, single-page PDF, image, and localized component artifacts.
- Build and query scoped multimodal representations with provenance.
- Run typed specialist workflows for project needs, measurement, engineering context, product context, calculation, reporting, and QA.
- Reconcile parallel results, validate output contracts, and preserve unsupported conclusions as unresolved.
- Return a Qualified Opportunity Report with evidence, calculation basis, warnings, state, ownership, and the next governed action.
Workflow system, not a file chatbot
A conversational surface may start or continue work, but the durable product behavior is the workflow: registered sources, addressable artifacts, tenant-scoped retrieval, validated records, observable attempts, resumable state, and human release gates.
Detail: specialist roles
| Role | Responsibility | Output |
|---|---|---|
| Project Analyst | Project metadata and typed need extraction | Need groups and project context |
| Plan Measurement Analyst | Scale, geometry, and measurement assessment | Measurement audits, status, assumptions, and questions |
| Engineering Agent | Rules, constraints, calculations, and technical fit | Engineering rationale and calculated quantities |
| Products Agent | Approved catalog and application matching | Candidate products and product evidence |
| Report Writer | Structured synthesis of validated artifacts | Report narrative and methodology |
| Action Strategist | Selective technical-commercial follow-up | Zero to three evidence-backed actions |
| Report Reviewer | Material gaps and unresolved decisions | Targeted drill-down prompts |
| Managed QA reviewers | Optional evidence and engineering QA supplements | Additional warnings or review findings |
Detail: seven distinct confidence signals
| Signal | Meaning | Boundary |
|---|---|---|
| Screening rank score | Orders candidate document sets for attention. | Pursue/review/reject recommendation-not a probability of winning. |
| Technical fit | Combines direct product, application, incumbent, semantic, and design signals. | Profile-relative screening component-not final engineering fit. |
| Pursuit readiness | Reflects commercial, quantity, evidence, and project metadata signals. | Commercial prioritization-not a bid decision. |
| Evidence confidence | Describes screening evidence diversity, text/page coverage, and semantic completion. | How inspectable the basis is-not proof that every conclusion is correct. |
| Pursuit confidence | Represents the strength of the deeper opportunity case. | Kept separate from product and quantity confidence. |
| Product-match confidence | Represents the support for a candidate product relationship. | Does not independently approve equivalency or suitability. |
| Quantity / measurement confidence | Represents source, scale, method, assumptions, cross-checks, and conflict support. | A separate estimating signal with its own verification state. |
Need help applying this page to an evaluation?Ask a product question