Three useful agents, and four that are not.
A field taxonomy from twelve months of production agent deployments across four industries. The list of things agents are not yet good at is longer than most vendors will admit.
Insights & Field Notes
On replacing manual QA with evaluation harnesses — and what it changes about how an enterprise engineering organisation is staffed, structured, and held accountable.
Read the field noteA field taxonomy from twelve months of production agent deployments across four industries. The list of things agents are not yet good at is longer than most vendors will admit.
What changes about a business case when the system delivering it is itself evolving on a weekly cadence. A framework for writing business cases that do not collapse at first contact with production.
The opening position. On why ATS Meridian is neither a consultancy nor a startup, and why that distinction matters more to the CIO than to the vendor pitching them.
A practitioner's guide to evaluation-first delivery — written for CIOs, CDOs, and the engineering leads who have to make AI work in environments where hallucinations are not acceptable and audit trails are mandatory.
Download the paper