I have trouble coming up with accurate estimates, leading to missed deadlines.

First, I recommend this Q&A: "How do I better plan my work during sprints?"

Underestimating the amount of time it takes to fully complete a task is a common failure mode I've seen among junior and mid-level engineers. It's important to remember there's much more to a meaty task (i.e. >2 weeks to complete) besides just coding:

1. Planning/technical design
2. Coding
3. Code review (this can take several iterations, and if your team isn't fast at code review, it can take 1 week+)
4. Landing the code and having it get to production
5. Monitoring the release in production and fixing bugs as necessary (this could include experimentation)

There are more of these steps as companies get bigger and at Series C, I imagine there's a good amount more process than "Just throw the code out there".

One aspect to this is that I take a lot of iterations before I fully understand something...

This is something I talk about in-depth in this other Q&A: "How do you get more independent?". In a nutshell, try to ask a bunch of questions (literally every single one you can come up with) all in 1 go as opposed to dragging it out over a long period of time. This will require some planning on your end before jumping into knowledge seeking as you try to figure out all the questions you have. And of course, write them all down after you do this exercise!

One strategy could be -

1. When listing down every smaller activity on a task level that will be adding up in that whole story.
2. u can have a template
3. just add hours for every task.
4. have some buffer
5. convert that hours to story points if you follow the above steps its hard to over commit . hope it helps