One, I think it is not fit for Hindi language podcasts. It recognised some words being spoken as suggested removable sounds on a consistent basis. So, i assume it cant recognize them as proper hindi words. In fact, they were proper hindi words. I cant say that this app is completely useless for hindi podcasters, but i hope it will get better down the line.

Two, Enahance audio feature seems really good. (Felt like cherry on the cake) It could remove vehicle sounds from nearby traffic, Room's AC humming sound. However, it can do even better if it can remove click sounds, muffles and ruffles nearby. Even if it identifies them as suggested removable sounds, we can review it and choose to keep or delete.

Three, Coudn't see any option to remove some portion of the clip manually. There were long silences in the audio which I was editing and wanted to remove those parts, but cud not see any option to remove the portion i did not want. And the algorithm could not recognise the long silences or gaps, where some silences were almost 2 seconds or more.

Four, the method to play relevant part where suggested removable part is there, seemed very innovative. However, give us key board shortcuts to play, keep or delete. As these three things are going to be used almost all the time while using this app. So, if you provide three shortcuts to users, i think it will cover 80% of their usage in the app.

Five, Couldn't edit the file name after it is imported in the editor studio.

Six, Export happens in Zip file, Why? Why it can't be just mp3/aac/wav directly

Seven, I think it doesnt have any option to create a folder structure to arrange files. However, it can be ignored if you are given only 15 days of file storage

Suggestions - The data should be trainable. We should be able to feed some portion of the clip to train the model to recognise them as removable sounds. Based on the feed, it should recognise similar sounds in the audio. This should be easy to incorporate in the code, i think.