0% read on this device
Browse the curriculum

Start Here

Vision

First Principles

Operating Model

Domain Model

Agent Factory

Runtime Architecture

AI Engineering

Autonomous Workflows

Verification & Delivery

Factory Platform

Quality Engineering

Security & Governance

Case Studies

Labs

Interview Practice

Research Journal

Reference

Curriculum/AI Engineering/A focused view of boundaries, contracts, state, authority, failure paths, and tradeoffs drawn from this chapter.
AI Engineering8 min readchapterQuick Read

Data, Knowledge, Context, and Semantic Engineering

Separate data fitness, reusable knowledge, semantic meaning, and task specific context so failures can be located and corrected.

Status: Draft for studyRisk: highLifecycle: plan · execute · verify · learnContent reviewed 2026-08-30Maturity guide →
Claim boundaryThis is curriculum guidance. It does not by itself prove a production implementation.
architecture mode

A focused view of boundaries, contracts, state, authority, failure paths, and tradeoffs drawn from this chapter.

Whiteboard exercise

Reconstruct and defend this chapter’s architecture.

Reconstruct the architecture, name each boundary, and defend the tradeoffs.

agent runtimemodelscontext8 min chapter
Open the source exercise

Design a knowledge and context path for a multi-repository security change. Include source registration, missing-data checks, lexical and vector retrieval, entity resolution, reranking, permission filters, contradiction handling, context budgeting, citations, and four distinct evaluation layers. Introduce an obsolete policy document that ranks highly and show how the system detects it.

4. Tradeoffs and alternatives

Centralizing knowledge simplifies governance and discovery but can create a stale copy of systems that already have authoritative APIs. Live retrieval preserves currentness but increases latency and dependency risk. A hybrid approach may index discovery metadata while resolving consequential facts from the source at decision time.

Fine-grained chunking improves targeted retrieval but can remove necessary context. Large chunks preserve narrative structure but consume budget and blur ranking. Knowledge graphs improve explicit traversal and lineage at the cost of modeling, ingestion, and consistency work.

An ontology is valuable when several sources repeatedly disagree about meaning. It is premature when a small controlled vocabulary and stable identifiers solve the actual problem. Semantic engineering should remove measured ambiguity, not create a speculative enterprise model of everything.

5. Current Mission Control Implementation

At study commit d902fae, Mission Control has provenance-backed retrieval, graph relationships, planning, Attempt-bound Context Packages, context evaluations, configuration drift scans, versioned context manifests, and content-hash checks. Factory Memory is advisory and cannot satisfy acceptance.

These mechanisms provide a strong context-governance foundation. The studied evidence does not establish a complete production source registry, connector and checkpoint lifecycle, data-quality gate, semantic-contract registry, permission-aware hybrid retrieval service, or independently benchmarked reranking pipeline. The graph and memory mechanisms should therefore not be presented as a general enterprise knowledge system.

External review

Review this chapter.

Challenge a claim, boundary, missing failure mode, unclear term, or unsupported evidence statement.

  • Claim
  • Boundary
  • Failure
  • Evidence