FactVerse Modules

FactVerse modules package platform capabilities for repeatable industrial workflows. The documentation structure separates base platform capabilities from module-specific operational interfaces.

Prerequisites

Before selecting a module, confirm:

• tenant, site, and operating owner are known;
• source systems and data owners are identified;
• asset, equipment, scene, line, cleanroom, or work-record IDs can be mapped;
• read-only review can be tested before compute or write actions;
• MCP scopes and product permissions match the workflow boundary;
• a review record will capture evidence, assumptions, decisions, and feedback.

Module selection flow

Base platform

The base platform provides shared context, governed access, common computation, and approved action patterns.

Data foundation

Data Fusion Services is the operational data layer used to connect source systems, map source values to operational targets, review data quality, and prepare governed datasets for Agent, Inspector, BI, and simulation workflows.

Start with Data Fusion Services when a module workflow depends on source-system signals, work records, meter readings, inspection records, or fused datasets.

Industry modules

Module documentation starts with MCP integration and reference pages that are stable enough for customer-facing use. The current code-evidence scope covers facility operations, predictive maintenance, Physical AI, and SemiOps.

Start with Facility Operations for SmartFM and Inspector operations, Predictive Maintenance Operations for the platform maintenance capability, Physical AI for simulation-ready asset and scenario workflows, or SemiOps Cleanroom and Facility Operations for semiconductor cleanroom and facility workflows.

Each module now has a matching AI tools page:

Module	AI tools page
Facility Operations	Facility Operations AI Tools
Predictive Maintenance	Predictive Maintenance AI Tools
Physical AI	Physical AI Tools
SemiOps	SemiOps AI Tools

Source data inputs

Module workflows usually need a prepared data package:

Input	Use
Asset or equipment identity	Connects UI records, source systems, Agent tools, and review output.
Source-system records	Provides signals, work orders, inspections, documents, utility readings, or simulation constraints.
Data quality notes	Shows stale values, missing fields, rejected rows, and review limits.
Permission and scope plan	Defines what the UI, backend, MCP client, or Agent can read, compute, or draft.
Review owner	Accepts or rejects output before it affects operations.

How to use module docs

Start with the MCP integration guide, then review scope reference for the module you are integrating. Tool availability should be discovered through MCP at runtime and cross-checked against the MCP Tool Reference.

Expected output

After reading a module guide, an implementation team should know:

• which user workflow the module supports;
• which source data must be prepared;
• which UI routes, APIs, MCP endpoints, or AI Engine endpoints are relevant;
• which outputs require human review;
• where to record evidence, limitations, and feedback.

Validation checklist

• Module choice matches the operating workflow.
• Data preparation path is clear.
• Runtime tool discovery matches the documented endpoint and scope plan.
• AI-assisted or computed output is reviewed by an accountable owner.
• Write actions stay as drafts until the approval path is configured.

Failure handling

Problem	Response
Module boundary is unclear	Start from the operating object and owner, then choose the smallest module that covers the task.
Source data is not ready	Use DFS connector, mapping, quality, dataset, or recipe pages before exposing the workflow.
Tool is unavailable at runtime	Check module enablement, endpoint, API key, and scopes.
Output cannot be approved	Keep the result as a review note and record missing evidence or owner decision.