The problem.
Every transparency law in Europe has the same failure mode: it works exactly once, for the one journalist who knows the right form. Romania’s Law 544 / 2001 is no exception. The right to request a record exists; the ergonomics of fifteen thousand such requests, against five thousand public bodies, with proper tracking, response parsing, and downstream routing into investigative journalism — that’s a software problem the law doesn’t solve for you.
What I built.
ACTFOIA is the methodology. It assumes Law 544 is the floor, automates the surface above it, and exits cleanly into investigation pipelines. The components are deliberately separable — most civic groups don’t need the whole stack, they need three of the modules and the legal scaffolding.
Specimen request
(Auto-generated by ACTFOIA · template L544-PROCUREMENT-v3)
The pipeline.
Where the wins came from.
Three places, in order of impact: (1) the template library — ~80% of requests against ~80% of bodies fit one of fourteen domain templates, which collapses the per-request human time from hours to seconds; (2) the response parser — most Romanian public bodies reply with a scanned PDF, and OCR + structural heuristics turn that back into rows; (3) the escalation timer — most non-compliance is sorted by sending the second letter on day 11, which the manual workflow never does.
Where it goes next.
The talk I gave at HOPE XV (2024) was the public release of the methodology with enough technical detail for civic groups in other jurisdictions to fork it. The statute differs by country; the methodology doesn’t. If your country has a transparency law and a journalism community that uses it, ACTFOIA ports.
What I learned.
The single highest-ROI move in a FOIA pipeline isn’t the legal posture — it’s the parser. We spent two years writing better request copy before we realized that the bottleneck was always on the inbound side. The day we shipped a working PDF-OCR pass, time-to-story for journalists dropped from weeks to days.