The first year of Inora in care.
What care professionals used the first version of Inora for, when, and where it fell short. This is the outline of the study; there are no results yet.
Founder and CEO
Outline of a studyDraft
Draft
This is an outline. It deliberately holds no result. We publish findings only once the organisations have agreed and clinical readers have graded the answers and draft records blind, so that no figure about the first version’s failures leaves us without that grading.
1
The question.
What did care professionals ask the first version of Inora, in which shifts and with what work in mind, and where did the answer not help?
2
The method.
What we count and why: the units (a conversation, a sitting), the exclusions, how conversations were labelled and by whom, and where people checked the models’ labels.
3
The work.
Which kinds of work the first version did: looking things up, writing, thinking along about a client, trying it out. And how that shifted over the year.
4
Inside a conversation.
What happens between the first question and the last answer: how often there is only one question, what people have rewritten, and what a poor first answer does to the rest.
5
Use over time.
How a habit forms, or does not, by role and by time of day.
6
Where the first version fell short.
What went wrong, with, for each point, how we established it and what the next version does differently. This part opens only after the blind grading.
7
Limits of the study.
What this study cannot say: it describes the first version, which is no longer in use, and gives no verdict on the current one.