

A graphical display associated with a case
Its background section names, in 2015, the failure that limits enterprise AI adoption today: analysts cannot later recall the specific information they relied on, and operators who cannot see the reasoning behind a recommendation show non-compliance.
The claims make grounding structural — evidence, interpretation and recommendation are first-class fields bound to a case, not commentary attached afterwards. The interpretation and the recommendation are written at a central monitoring center; they are read at the operating site, by the person who has to act.
Figures 3 and 4 of WO2017048218A1 — the case data structure, with evidence, interpretation and recommendation as fields; and the same structure bound to a strip of widgets.
This is the earliest of four instances of one architecture across eleven years. Everything I have built since about explainability was already structural here.
Continued in Asset Performance Management (2014–2015) and KBOps (2022).
Official title: Method and apparatus for providing a graphical display associated with a case. Filed with David Sean Farrell.
One international application · filed Sep 2015 · published Mar 2017
















