Glages: What the Factory Is and Is Not

New products are often described through familiar software categories. Glages may be compared with agent frameworks, workflow engines, rules systems, knowledge graphs, retrieval products, and evaluation tools because each of them touches part of the problem.

The comparison is useful only after the main category is clear.

Glages is a factory for producing reusable formal models of business processes and business domains for AI automation.

The factory may use several existing technologies, but none of them defines the product.

Not an Agent Framework

An agent framework helps an agent use tools, memory, plans, and external systems.

Glages defines the activity before the agent acts. The agent receives approved typed Input and operates only inside modeled choices and boundaries.

An agent framework may execute inside a Glages-based system. It does not replace the formal model.

Not a Workflow Engine

A workflow engine coordinates steps and transitions. It can be useful for execution, especially when the process is already defined.

Glages addresses the prior question: what do the steps mean, which Entities and facts participate, which changes are valid, which results exist, and where execution must stop?

A workflow may implement part of a model. The workflow itself is not necessarily the complete business model.

Not a Rules Engine

A rules engine evaluates conditions and produces decisions. Rules are important, but a business process also requires Entities, states, relationships, actions, changes, results, evidence, and process boundaries.

Glages places rules inside a broader formal model of the activity.

Not an Ontology or Knowledge Graph

Ontologies and knowledge graphs can describe concepts and relationships. They may help organize or retrieve knowledge.

Glages models executable business activity. It must define not only what things are related, but what may happen, under which facts and conditions, through which permitted actions, with which changes and results.

Graph structures may participate in implementation or analysis. They do not by themselves constitute controlled automation.

Not a Retrieval Product

Retrieval gives an LLM relevant source material. It can improve the information available during interpretation.

Retrieved text remains text. It may conflict, omit required facts, or require judgment. Glages converts approved meaning into a formal model so that the complete process does not have to be reconstructed from retrieved passages during every execution.

Not an Evaluation or Observability Layer

Evaluation and observability help measure what an AI system produced and how it behaved.

Glages defines what the system was permitted to do before execution. Trace and verification remain important, but they operate against the formal model rather than only against examples or subjective scoring.

Not Merely a C# System

Glages is built directly on C#, and the typed C# model is the technical source of truth. Compilers, analyzers, generators, runtime validation, and tests support verification.

Those implementation choices explain how the factory works. They do not define the market category.

A buyer is not purchasing a programming language. The buyer is obtaining reusable formal models and the ability to connect them to products and systems.

The Positive Definition

Glages produces the missing operational structure that agents, programs, interfaces, and services can use.

It creates reusable process models, composes them into business-domain models, connects them to companies, verifies them, and maintains them in a growing model library.

Other technologies may participate at different stages. The formal model remains the source of permitted meaning.

That is what the factory is.

Complete-Series Consistency Note

The seventeen articles in this file use one shared product hierarchy:

base concepts and relationships
→ reusable process models
→ composition of processes
→ composed business-domain model
→ company connection

They preserve the following meanings throughout:

• Glages is a factory for producing reusable formal models of business processes and business domains for AI automation.
• An automation model is formal and machine-checkable.
• The model is the automation.
• No model means no automation.
• LLM output is a candidate until independently verified.
• The agent operates only inside approved typed Input.
• Reusable models remain separate from company connection.
• Shipment and delivery are separate processes.
• Business results remain distinct from technical failures.
• Public explanations do not disclose the detailed production recipe.

Review status: Approved as a combined first complete draft after cross-article consistency review. Each article may still receive ordinary editorial polishing before separate publication, but no unresolved conceptual decision is intentionally left inside this file.