Hyunjoong Kim's Technology Blog
Practical writing on AI agents, SDLC documentation automation, full-stack architecture, and product validation.
31 results
Agent Governance Is the Architecture: The Real Control Plane in the MCP Era
If MCP standardized tool connectivity, permissions, policy, audit, and isolation are what make production AI agents safe.
MCP Is Not the Architecture: What Production AI Agents Actually Need
MCP, multi-agent workflows, and tool integrations are hot right now, but production systems live or die by context, state, and control planes — not the protocol alone.
Agentic AI Needs a Control Plane, Not Just Prompts
Google Workspace's agentic expansion and NVIDIA's BlueField-4 STX point to the same shift: AI is becoming an operations-architecture problem. This post reframes AI workloads through a control-plane lens.
Agentic Workflow Architecture: How to Run AI Agents as a Product
Model quality matters, but execution structure matters more. This post lays out a practical framework for designing agentic workflows around decomposition, routing, state, and observability.
AI Agent Orchestration Is the Real Bottleneck: Design These Layers First
Agent systems usually fail in orchestration before they fail in model quality. This post breaks down workflow design, tool permissions, fallback, and evaluation as architecture problems.
The Real Value of AI Coding Agents in 2026: Beyond Autocomplete to Review, Tests, and CI Recovery
AI coding agents are moving from code completion toward pipeline automation across reviews, test generation, and CI recovery.