Yigit Tas·Jul 15Your Prompts Have No Test Coverage314 green tests, zero tests on the file that steers the agent. So I removed my rules one by one and measured what changed.
Yigit Tas·Jul 7In Agentic AI, a Status Is a MoodThe fifth and final post of the series: what building agentic systems taught me about the difference between a claim and a record.
Yigit Tas·Jul 1In Agentic AI, the Output Is Not the EvidenceAn agent producing something and a system leaving behind an auditable trail are not the same thing.A response icon1A response icon1
Yigit Tas·Jun 19In Agentic AI, Success Is Not EvidenceIn agentic AI, a “success” is just a claim. Trust is something a signal has to earn.A response icon1A response icon1
Yigit Tas·Jun 9In Agentic AI, Orchestration Isn’t Enough: You Also Need to Define the Role of DataBetter decisions come from clear information roles, not from putting everything into context.
Yigit Tas·Apr 2Agentic AI Is More Orchestration Than ModelWhat makes agentic systems useful is not the model alone, but the orchestration around how decisions are produced, updated, and controlled.
InhepsiburadatechbyYigit Tas·Oct 12, 2023Performance Testing and k6Performance testing is a test methodology that evaluates the responsiveness and stability of a server, network, or application under a…