Explicitly use the Note as the source of letter generation
complete
J
Jamie McConnell
Heidi's processing can be thought of as mutiple layers of indirection.
Voice is processed to make a transcript
The transcript is processed to extract concepts
The concepts are transformed into a note or letter
The last stage is the problem. The interface presents the note first, with a button inviting the user to convert it to a letter. This suggests a workflow in which the notes are being used as the basis for the letter, but this is not the case.
In practice, the notes and the letter are very obviously generated from the same data, and will contain the same errors, even if those errors have been corrected in the notes section.
This is bad for several reasons.
Firstly, the UI suggests that the note will be used to generate the letter (you may not have intended this, but that is the impression that is given), so there is a risk of user error there.
Secondly, once the user understands what is really happening, there is an obvious duplication of effort, where mistakes have to be corrected in both the note and the letter (and if I wish to produce e.g. onward referral letters, then I need to correct those too).
Thirdly, errors are easy to correct in a bullet-point list, but are harder to correct in a narrative letter (simply due to the additional cognitive burden of reading prose vs notes).
The obvious solution is for Heidi to work the way that the interface suggests is actually happening:
Audio > Transcript > Concepts > Notes > Letters
For bonus points, have the letters automatically mark themselves as containing outdated information, if the Notes are changed.
H
H Jalal
I don't think this is resolved...?
I understand that the Audio > Transcript is the first layer.
Then Transcript + Context > Notes. Now, in generating "Notes", Heidi prompts me which template to use - and I would say it gets 75% of the structure of that template right.
BUT.
I now have to manually click Create and re-select that same template again to get a more accurate and complete note nearing 100% of the template instructions. For me, this is the Notes > Letter additional step.
And to this post's point, any changes I make in the 'Notes' does NOT get reflected in the final letter - to get Heidi to implement any updates I have observed 3 options :
- Resume recording and state the new information - hit and miss, sometimes Heidi will ignore this additional info, and other times it will incorporate it
- Type it in context and ask it to regenerate/sync - again hit and miss but mostly gets it
- Use the Ask Heidi feature and modify parts of the note - clunky, least preferred, and adds a lot of admin time to what should be a very smooth, streamlined experience.
J
Jamie McConnell
A welcome change, but it’s not obvious from the user interface that this is what’s happening. The workflow has become a total black box now.
Mo from Heidi
complete
Given the main contention of this ticket is resolved, I'll mark this ticket as complete.
Mo from Heidi
Hi Jamie McConnell - I totally forgot to circle back and found this serendipitously on leisurely scrolling submissions. We released last Friday precisely what you're describing here. Artefacts (letters, documents, anything non-note) are produced secondary to the core note itself - we find this produces vastly superior non-note artefacts.
Regarding the other things you and Callum Lawson have mentioned here:
- Mark non-note artefacts as outdated when notes changed: I've logged this in our backlog.
- Persistence of edits when regenerating notes: We're actively working this problem right now. It's tricky - we're juggling a couple of technical approaches that have emerged as possible winners. I do like a "this letter is outdated" as a visual stopgap until we can build a true and effective sync engine...
Tagging Theo from Heidi for visibility
J
Jamie McConnell
Mo from Heidi The clear visual indication is the key here. Syncing would actually be counterproductive without it.
Mo from Heidi
Jamie McConnell Hi Jamie - letting you know point 2 of my comment is done. Edits will persist when regenerating notes.
C
Callum Lawson
This! The fact that edits are lost when regenerating notes is a problem in and of itself.
Sometimes I'll start editing the mistakes Heidi has made in the note, then realise it's too much/too little detail, change the detail level and have to make all the edits again.
Heidi should track edits to the generated casenote against the transcript it was generated from and use those edits in any future generation (kind of like the "Additions" function but post-transcription)