AI Tutorials
August 29, 2026•18 min read
AI Agent Loops and Graphs: A Production Field Guide
Turn agent loops, graph orchestration, scoped retries, evidence gates, and human approval into one testable production architecture.
Blog
Insights, tutorials, and updates from the ZharfAI team
AI Tutorials
Learn about AI technologies and implementations
AI Tutorials
August 29, 2026•14 min read
How to Build AI Decision Explanations That Can Be Tested and Challenged
A practical evidence contract for producing faithful, useful AI decision explanations—and rejecting plausible rationales that cannot be proved.
AI Tutorials
August 29, 2026•16 min read
Six Businesses, One AI Desk: Grok Bot + Kimi K3 Ledger Architecture
A source-checked build guide for six ledger-owning agents, one routing desk, evidence-bound handoffs, and human control over consequential actions.
AI Tutorials
August 28, 2026•14 min read
One Agent Cannot Check Itself: Separation of Duties for AI
A control architecture for separating proposal, approval, execution, and reconciliation when AI can trigger consequential business actions.
AI Tutorials
August 27, 2026•14 min read
The Safe Change Unit: Releasing AI Without Changing Everything at Once
A practical change-control architecture for deciding what must be versioned, tested, canaried, and restored together when an AI system evolves.
AI Tutorials
August 26, 2026•13 min read
Permission to Read Is Not Permission to Release: An AI Output Gateway
A practical architecture for deciding when AI-generated content may leave, must be minimized, needs human review, or must be blocked.
AI Tutorials
August 25, 2026•13 min read
Do Not Mix the Maps: A Safe Migration for Embedding Indexes
A field guide to rebuilding, evaluating, cutting over, and rolling back vector search when the embedding model, chunks, or index must change.
AI Tutorials
August 24, 2026•15 min read
A Number Is Not a Quantity: The Measurement Contract for AI
A field guide to units, currency, precision, rounding, and deterministic calculation before AI-generated numbers can create business effects.
AI Tutorials
August 23, 2026•14 min read
The Last Release: Retire an AI System Without Leaving It Running
A field guide to closing intake, authority, dependencies, work, evidence, and data so an AI service can be replaced or retired with a verifiable end state.
AI Tutorials
August 22, 2026•13 min read
The Answer Is Optional: An Abstention Contract for AI
A practical guide to deciding when AI should answer, seek evidence, defer, or refuse using calibrated signals, risk–coverage curves, and fallback capacity.
AI Tutorials
August 21, 2026•14 min read
Before You Install the Tool: A Supply-Chain Gate for AI Agents
A field guide to admitting, quarantining, or rejecting MCP servers, plugins, and agent tools using provenance, capability tests, and enforceable runtime limits.
AI Tutorials
August 20, 2026•13 min read
Drift Is a Clue, Not a Verdict: Monitoring AI After Deployment
A field guide to baselines, delayed outcomes, change attribution, and deciding when deployed AI should be watched, constrained, rolled back, or rebuilt.