at a game studio of 300 i would wander amd chat with the various non-coder teams and just watch their workflows. for the tech artists a big blocker was gui access to enums in shaders (custom engine). i just set several of them up to compile the engine and taught tyem how to link enums to the editor gui, saved them many weeks of wait time. saw the lighting bake loop would take 1 year to run fully and a local test loop was 8 hours. asked my boss for time and rigged up a cluster of 64 ps4 dev kits and brought that down to 45 minutes, going from less than one local iteration a day to 9, and less than one a year to 3 a week for full bakes (cluster would bake when it wasnt helping am artist).
at super large mega corp i do similar, constantly meeting people for coffee and understanding more than just my cog. Eventually I got good enough at it and a reputation for improving things that I just get pushed/pulled at problem/opportunities areas, but you have to work up to it.
while none of this is doing a startup, if youre not curious, youll never see the myriad of things that could be improved with software, nor will many non tech folks, or really anyone who doesnt think "outside the box" as the olds say. challenge for devs is understanding where that tech is worth it and how to sell it, and not become that angry dev who is just yelling at everyone and politics for not taking every one of their half baked ideas as words from on high.
An example from my day to day in a large corporation: we have to enter time weekly: how much time did you spend on project X, Y or in training/meetings/etc. Sure there is software that does this, but it's terrible There are many issues with it that should be trivially solvable.
Now, I will admit that time tracking software is a huge field with lots of strong competition, so this may not be a field you'd want to enter, but it's an example of the problems you would come across during an average workday that are ripe for better solutions.
Try meeting people not in your field. Yes, it will take effort. Yes, it will be uncomfortable. Because of those challenges, fewer people have explored that space - so there are still fertile grounds for new products.