I'm pretty open, although we still haven't achieved any success yet... very possibly because of the mistakes I've made! So probably there's no reason anybody would care about anything I have to say.
Anyway, I'd say the single biggest mistake I've made over the years is being to quick to ignore the customer development process, and start building stuff without getting enough feedback from prospective customers. I've spent too much time "heads down" writing code and not enough out in the field talking to people.
Second, and related, mistake has been being too ambitious from a product standpoint. Or to flip it around, not being focused enough. That is, I've been too quick to start adding more features to the product(s), as opposed to taking a hyper-focused approach, picking a subset of features to truly finish, including all "fit and polish" stuff. So far "fit and polish" isn't something we've been very good at.
Another decision that I don't consider a mistake, but which is arguably contestable, is choosing to go the bootstrapped / self-funded approach, instead of going out and pursuing outside money. I'm happy with this, but it has meant continually being pretty resource constrained. We can't yet, for example, hire employees. So all the work is done by the founders, all of whom still work a day-job so far. You could argue that if we had raised some money, we could have moved a lot faster. OTOH, you could at least speculate that if we'd raised some money, we'd be out of business now. The way things are, unless one of the founders dies or quits, we're almost 100% certain to not go out of business. We can manage this burn rate almost indefinitely. The onus on us now is to get out there, engage with prospects, and get something out the door that somebody will pay for.
That said, I think we're close, especially with this neuralobjects.com thing.