Decisions & Best Practices
Decision frameworks and architectural best practices for enterprise agentic AI — from implementation pattern selection to prompt engineering and data governance.
Decision Frameworks
The most common architectural questions in enterprise agentic AI implementations don't have universal answers — they depend on the specific use case, governance requirements, and organizational context. These frameworks guide you to the right answer for your situation.
Agent vs. MCP vs. Orchestration
The primary decision framework: when to use an agent, when to expose via MCP, and when to use a recipe.
MCP Server vs. MCP App
When to build a server (exposes capabilities) vs. an app (consumes capabilities via structured UI).
Best Practices
Prompt Engineering
Best practices for writing reliable agent prompts at each of the four levels.
Data Governance
How to govern data access, retention, and usage in enterprise AI agent deployments.
Security
Security practices for enterprise AI agent development and deployment.
AI Model Selection
How to choose the right AI model for your enterprise agent use case.