Pivoting from investment banking

I spent the first two years of college on the investment banking recruiting track. Networking calls, cold emailing, technical prep, the BIWS 400 IB interview questions, Wall Street Prep courses, case competitions, multiple internships, staying active in finance and investment banking clubs, and maintaining a high GPA. By junior year I was in active interview prep, with CFA prep running alongside it.

During that year I proposed a side project to two of my friends. Given everything else on my schedule, work on it was limited to one hour calls at night. One of the friends involved had previously worked at Citadel and Meta and was at a startup at the time.

About two weeks in, he asked whether I actually wanted to do investment banking. I did not have a clear answer. I had picked the target early and had not revisited it since, and most of what kept me on it was the prestige attached to the outcome rather than the work itself. He said he thought I would do better in a startup environment, and that if I was interested he would talk to the founder about bringing me on.

The decision carried a real cost. Accepting meant dropping CFA prep and interview prep, and canceling the interviews already on my calendar. It also meant trading two years of accumulated progress for a position where a return offer looked nearly impossible. The company was working on frontier 3D technology and was deeply technical, and at that point I had been coding at night for two weeks. Nevertheless, I decided to take this rare opportunity, ultimately betting on myself.

He put in a word, and a meeting with the founder was set for a few days out. I spent those days building projects with their public library, working on a 4D livestream that ran on a voxel encoder. When I showed the founder what I had made, he was impressed with how much had come together so quickly, and assumed my friend had helped me with it. After a short conversation about my projects and what I was looking for in working with the company, he offered to have me work onsite with the team for the day.

Before any of it started, the friend who made the intro told me what to expect. I would be treated as a complete beginner, given self-contained work to start, and eventually some easy frontend tasks. What I got was completely different. I was handed full access to the codebase right away, which surprised me, and the founder had me ship a few features on the company website that same day, front end and back end. I completed the work. The following day, a Friday, he moved me onto what the team was actively building. The task was building a Gaussian splat, which meant modeling the scene myself in Blender, rendering out views of it to train from, learning Modal to run the training on GPUs, and then wiring up the splatting library to render the result. I spent the weekend on it and had it finished before I came in on Monday. The tasks kept coming after that, with the difficulty increasing each time.

About a week from my start date, he decided I had shown enough agency and capability and shifted from assigning me tasks to having me own a project end to end. The project being building a 4D livestream on a voxel encoder, something that would have been difficult even for the experienced members of the team. He told me he had originally planned to start me on something much easier, and a week of watching me work changed his mind. I shipped it in about two weeks. That was what ultimately secured an official internship, less the meeting or the demos beforehand than how far the work got in a short window.

From there the scope only grew. I shipped multiple projects end to end, covering backend, frontend, design (product, UI/UX), and marketing/GTM, even helping craft pitch decks and draft investor updates, where my finance background finally got its use. I finished my degree while working full-time and commuting to Brooklyn, and the internship converted to a full-time offer.