The one thing that corporations are extremely good at is destroying all the initiative and add so much bureaucracy to the development that you loose interest to improve / inovate as you are paifully aware that this will bring in more documents, jira tickets, endless "bicycle shed" discussions, endless covincing of managers that only care about their schedule (which you will corupt by "unknowns"), endless politics as there is always some architect/team lead that "must be the smartest" (and typically those comming to those positions are excelent in navigating buereaucracy and politics and are rarly inovators as those endanger manager schedules, adding risks, ...) and so on.
Too many cooks spoil the broth.
At the end you just figure out it is not worth it , do what you are paid for and you inovate at home.
If those inovations become a startup you just leave.