oniru.ai8 min read
Build an LLM step that returns structured JSON you can trust.
A model in the middle of a workflow returns whatever it decided to write. Build the step so the node after it always receives the shape it was expecting.
Notes from systems in production.
Practical writing on automation delivery, APIs and the operational details that make systems reliable.
Showing notes tagged AI.
Articles
oniru.ai7 min read
Build a document extraction pipeline with a confidence gate.
A pipeline that extracts fields at ninety percent accuracy and writes all of them has not saved anybody time. It has created a data problem for later.
oniru.ai7 min read
What clients actually mean when they ask for AI.
A request for AI is a budget and an expectation, not a requirement. Behind almost every one there is a specific, boring, buildable job worth finding first.
oniru.ai5 min read
The EU AI Act transparency rules, and what agencies must do.
Enforcement began on 2 August 2026 and the rules reach further into ordinary agency work than the coverage suggested. What is in scope, and what is not.
Have a client asking for something you can’t easily build?
Send us the requirement. We’ll tell you how we’d approach it, what it would require and whether it makes sense to build.
No pitch deck required. A rough description is enough.
Or write to [email protected]