I spent two years building an automated system for building resident block schedules. I made it to a beta with a very large residency, but for a combination of technical and incentive reasons it was impractical to grow. I actually revisited the concept recently with a revised business model that might have addressed some of the growth challenges. I did a kickstarter, but didn't generate enough interest to pursue it again.
I also spent maybe 1 year prototyping a system meant to analyze the performance of wealth managers. I used financial statements from friends and family to see whether I could produce anything useful. But, the more I got into it, the more I realized it was difficult to produce a compelling automated analysis, even given a complete history of all the manager's transactions. It was too easy to swing the result by subtly changing the assumptions.
I also investigated an all-inclusive management system to help foundations for public high schools manage fund raising, etc. Again, I did a kickstarter-like campaign for it, and found inadequate demand.
I also spent maybe 1 year prototyping a system meant to analyze the performance of wealth managers. I used financial statements from friends and family to see whether I could produce anything useful. But, the more I got into it, the more I realized it was difficult to produce a compelling automated analysis, even given a complete history of all the manager's transactions. It was too easy to swing the result by subtly changing the assumptions.
I also investigated an all-inclusive management system to help foundations for public high schools manage fund raising, etc. Again, I did a kickstarter-like campaign for it, and found inadequate demand.