Ian Parent

building iris — the agent eval standard for mcp. writing about agent eval, cost visibility, and what happens when agents run in production.

CA, USA Joined on Mar 17, 2026 https://iris-eval.com

Education

Cal Poly (SLO), B.S. Mathematics

Work

Founder & Builder, Iris (iris-eval.com)

Ian Parent

Apr 14

Closing the Eval Gap: From Lenient Defaults to Signal That Matters

#mcp #aiagents #observability #opensource

5 min read

Ian Parent

Mar 29

Why On-Chain Agent Actions Need Pre-Flight Eval

#mcp #aiagents #observability #opensource

6 min read

Ian Parent

Mar 29

Output Quality Score: The Single Number That Tells You If Your Agent Is Good Enough

#mcp #aiagents #observability #opensource

5 min read

Ian Parent

Mar 27

Self-Calibrating Eval: The End of Manual Threshold Tuning

#mcp #aiagents #observability #opensource

5 min read

Ian Parent

Mar 24

The Eval Loop: Why Evals Are the Loss Function for Agent Quality

#mcp #aiagents #observability #opensource

6 min read

Ian Parent

Mar 24

Eval-Driven Development: Write the Rules Before the Prompt

#mcp #aiagents #observability #opensource

5 min read

Ian Parent

Mar 23

Eval Coverage: The Metric Your AI Agents Are Missing

#mcp #aiagents #observability #opensource

4 min read

Ian Parent

Mar 21

The Eval Gap: Why Your AI Demo Works and Production Doesn't

#mcp #aiagents #observability #opensource

4 min read

Ian Parent

Mar 21

Eval Drift: The Silent Quality Killer for AI Agents

#mcp #aiagents #observability #opensource

4 min read

Ian Parent

Mar 21

The AI Eval Tax: The Hidden Cost Every Agent Team Is Paying

#mcp #aiagents #observability #opensource

6 min read

Ian Parent

Mar 20

Agent Errors vs Application Errors: Why Your Error Tracker Can't See AI Failures

#mcp #aiagents #observability #opensource

6 min read

Ian Parent

Mar 19

MCP Meets OpenTelemetry: Bridging Agent Observability and Infrastructure Monitoring

#mcp #aiagents #observability #opensource

6 min read

Ian Parent

Mar 19

Toward an MCP Observability Specification

#mcp #aiagents #observability #opensource

8 min read

Ian Parent

Mar 19

How to Evaluate AI Agent Output Without Calling Another LLM

#mcp #aiagents #observability #opensource

7 min read

Ian Parent

Mar 19

MCP Observability is the New APM

#mcp #aiagents #observability #opensource

6 min read

Ian Parent

Mar 18

Heuristic vs Semantic Eval: When <1ms Matters More Than LLM-as-Judge

#mcp #aiagents #observability #opensource

7 min read

Ian Parent

Mar 18

The Cost of Invisible Agents: What $0.47 Per Query Looks Like at Scale

#mcp #aiagents #observability #opensource

6 min read

Ian Parent

Mar 18

Why Your AI Agents Need Observability

#mcp #aiagents #observability #opensource

5 min read

Ian Parent

Mar 18

Why Every MCP Agent Needs an Independent Observer

#mcp #aiagents #observability #opensource

6 min read

Cover image for The State of MCP Agent Observability (March 2026)

Ian Parent

Mar 17

The State of MCP Agent Observability (March 2026)

#mcp #aiagents #observability #opensource

9 min read

Forem

Ian Parent

Badges

Writing Debut

Skills/Languages

Currently learning

Currently hacking on

Available for

Closing the Eval Gap: From Lenient Defaults to Signal That Matters

Why On-Chain Agent Actions Need Pre-Flight Eval

Output Quality Score: The Single Number That Tells You If Your Agent Is Good Enough

Self-Calibrating Eval: The End of Manual Threshold Tuning

The Eval Loop: Why Evals Are the Loss Function for Agent Quality

Eval-Driven Development: Write the Rules Before the Prompt

Eval Coverage: The Metric Your AI Agents Are Missing

The Eval Gap: Why Your AI Demo Works and Production Doesn't

Eval Drift: The Silent Quality Killer for AI Agents

The AI Eval Tax: The Hidden Cost Every Agent Team Is Paying

Agent Errors vs Application Errors: Why Your Error Tracker Can't See AI Failures

MCP Meets OpenTelemetry: Bridging Agent Observability and Infrastructure Monitoring

Toward an MCP Observability Specification

How to Evaluate AI Agent Output Without Calling Another LLM

MCP Observability is the New APM

Heuristic vs Semantic Eval: When <1ms Matters More Than LLM-as-Judge

The Cost of Invisible Agents: What $0.47 Per Query Looks Like at Scale

Why Your AI Agents Need Observability

Why Every MCP Agent Needs an Independent Observer

The State of MCP Agent Observability (March 2026)