HomeAbout
About
An open framework for deciding whether a specific task should be automated — and, just as often, for arguing that it should not be.
The problem it was built for
Most automation calculators are lead-generation forms with arithmetic attached. They ask what you spend, assume AI removes most of it, and produce a number large enough to justify a call. The number is almost always computed by comparing your current output against a cheaper scenario that quietly produces less.
This does the opposite thing. It holds output constant, prices the supervision the work would actually need, and weights risk as heavily as cost — which means it returns “do not automate this” for a meaningful share of the tasks people most want to automate.
Who builds it
The framework is maintained by Nimrah Khan. The research behind the risk weightings, the leverage curve and the oversight contract is hers, and the parameter file carries her name on the citation line.
Why it is open
A verdict you cannot inspect is an opinion with a number attached. Every weighting, threshold and task score is published under CC BY 4.0, versioned, and changed only with a dated changelog entry naming what triggered it. You can reproduce any result on this site, disagree with a specific parameter, and cite the version you argued with.
How it is funded
The audit is free and stays free — no email, no account, no gate. Three things pay for it: fixed-price reports at $150, clearly labelled sponsored placements that sit outside the editorial sections, and nothing else. We do not sell editorial placement and we do not take commission on tools mentioned in reviews. If that changes, it will be said here first.
What it is not
It is not legal, financial or employment advice, it does not know your vendor quotes, and it does not price redundancy, retraining or the management attention a transition consumes. It is a starting point for a decision, and it tells you where its own edges are.