How does Bildstak unify twelve systems into one model?

Question

Accepted Answer

Bildstak unifies a project's systems through a federation layer that connects each source via its native API or file format, indexes every record by a shared project identifier — IFC GlobalID for BIM, activity IDs and cost codes otherwise — and exposes the unified model through a single query interface. A major construction project routinely runs across twelve or more specialist tools: an IFC authoring tool, Speckle or Autodesk APS for model sharing, Primavera P6 for scheduling, SAP or Maconomy for cost, Procore and Aconex for documents and RFIs, a BCF tool for coordination, IDS and COBie for standards compliance, Excel for quantity schedules and cash flows, and often a custom ERP or REST API for project-specific data.

Unifying these is a three-layer problem:

Layer 1 — Connection
Each source speaks a different protocol. Bildstak has native connectors for each: IFC file readers, P6 XER parsers, REST API clients for Procore/Aconex/SAP/Maconomy/Speckle/APS, BCF and COBie file parsers, spreadsheet readers, PDF and document indexers, and a generic REST API framework for anything else. Connection is configured once per project.

Layer 2 — Federation and joining
Once connected, records from each source are indexed by the project's join keys. For BIM projects, the IFC GlobalID is the primary key — every element in every source that references a model element by its GlobalID is joined to it. For non-BIM projects, P6 activity IDs, WBS codes and cost codes play the same role. The federation layer understands these relationships and can resolve a query that spans any combination of sources.

Layer 3 — Query and delivery
The unified model is exposed through a natural-language interface, purpose-built modules (Claims Builder, EVM, RFI register, Element Inspector, portfolio analytics, AI reports), and the autonomous Analyst. Every response cites its sources.

كيف توحّد Bildstak اثني عشر نظاماً في نموذج واحد؟

Related questions