I’ve been regularly testing and/or buying tools on AppSumo for some 5 years. I believe SoundMadeSeen may be among the most useful.

Until now, to make audiograms, I used the free plan of Headliner.app ("Make" tool). Quite handy BUT you can’t transcribe the audio and get the subtitles in the audiogram. You can in parallel use also on the free plan the "Eddy" tool to transcribe. With AI "Eddy" selects some excerpts/clips in the transcript, and enables you to send them to "Make" BUT : 
- you cannot yourself select the clip you want
- you cannot regenerate the AI selection
- you cannot send the whole transcript to "Make" to create a full-length audiogram with subtitles.

Thus, I was quite interested to see the SoundMadeSeen deal.

I have tested the tool somewhat thoroughly on one use case.

For sure, the UX/UI need further refinements. Some things still are somewhat crude or clunky.
However : 
- Rane, the developer, demonstrates great reactivity and openness to customer feedbacks
- he implemented a good dozen+ correctives and UX refinements in virtually no time, making the experience immediately better
- he did so while clearly replying and explaining each point
- he refunded credits I had spent in vain because of some UX problems since corrected

I think it is fair to label that as an excellent customer service, if not outstanding for now.

Thus, in say 4 days, in just one full cycle of testing, the app got better.

What’s more, I can say that the app is fully usable. You can see :

👉 the audiogram I created with SoundMadeSeen with subtitles : https://youtu.be/DCvfaoQqw6A (beware, the first 8 seconds are not subtitled)

👉 the blog post that SMS AI-generated from the audio and transcript of another (50 mn long) radio show. It was directly almost perfect, I just had to write an intro paragraph : https://mouvenfants.fr/libre-antenne-une-culture-du-silence-qui-perdure-face-aux-violences-faites-aux-enfants-2024-02

Testing the AI identification of relevant clips, I concluded that :
- the AI selected clips that were relevant and usable as such
- but also had missed other clips more relevant in my opinion

A significant caveat : test the app in your own working language. I spotted that in the first audio I processed, the transcription lacked 2-3 sentences spoken by a speaker that did not intervene elsewhere. The developer immediately went to the customer service of the transcription tool, Assembly AI. The problem lies there, and will not be solved at the soonest before next version "in a few weeks". Not a huge problem in the precise case where I was confronted to that, but worth mentioning. You may encounter a glitch of this sort with non-english transcription. For me, I count on the problem being soon resolved.

To sum up : this is a really good AppSumo deal, an already really handy tool, and still promising because has room for (easy or less easy) improvements.