Meaning Engine: deterministic computation over engineering knowledge graphs

I’m sharing Meaning Engine for technical feedback.

Meaning Engine is a deterministic computational substrate for typed engineering knowledge graphs.

It is intentionally NOT:

• an AI assistant
• an autonomous agent
• a general reasoning system

Instead, it focuses on reproducible computation over an explicit semantic graph.

The system produces:

• path traces
• directed boundary analysis
• rival explanations
• gap reports

The goal is to support engineering workflows where:

• reasoning must be inspectable
• claims must be bounded
• outputs must be reproducible

---

I’m not looking for promotion.

I’m looking for technical feedback from engineers.

In particular:

• Where is the architecture unclear?
• Where do we overclaim or underclaim?
• What objections would you raise?
• Where are the real limits?

---

Repository:
https://github.com/utemix-org/meaning-engine

Structured feedback thread:
https://github.com/utemix-org/meaning-engine/issues/32