Who am I? Recently promoted SDE2 at Amazon (retail side) with just under 2 YoE. Before that, I got a CS degree from a state school. What am I trying to do? Get a job as a quantitative developer or quantitative trader at a top buy-side trading firm. Why am I trying to do this? I have written mundane business logic for the past 2 years, and I am tired of it. I was interested in trading during college but was too focused on kickstarting my SWE career (had no CS experience before university) to invest much time into it. I want to try breaking in now that (1) I am financially stable and (2) I am still young (23). What am I looking for? Someone who is as committed as I am to this endeavor and in a similar stage of life (i.e. post-college and working a full-time SDE job). Since this background already makes us outliers in the quant world, it’d be nice to have another person to share the ride with. What is my quantitative background? No competition math experience. 1 month of programming contest experience. 200 total LeetCode problems solved over college (rusty now). Took discrete math, algorithms, combinatorics & graph theory, and number theory along with the usual calc I/II/III, linear algebra sequence. Retained the mathematical mindset but forgot all the key results / theorems. How will I do this? I will follow a 4-phase approach. 1. Improve algorithmic problem solving [1 year] - Current baseline: 1100 rating on codeforces.com - Core goal: stable 2000 rating - Stretch goal: stable 2200 rating 2. Build mathematical foundation [1 year] - Complete Art of Problem Solving curriculum - Complete Coding The Matrix book - Complete Analysis Vol. I by Terence Tao - Maintain stable 2000 rating on CodeForces 3. Build applied mathematics toolbox [2 years] - Quit job - Enroll in local state university as a non-degree student - Take courses that cover the equivalent of Volumes I/II/III of the Foundations Of Applied Mathematics series (https://foundations-of-applied-mathematics.github.io/) - Maintain a stable 2000 rating on CodeForces 4. Prepare for interview [6 months] - Solve all past AHSME problems from 1950 - current - Solve all past AMC 12 problems from 2000 - current - Solve all past AIME problems from 1983 - current - Maintain stable 2000 rating on CodeForces - Get to hard level on TraderTest Why this approach? The advice is sourced from an undergrad friend, who joined the industry out of college (quant dev at a strong prop shop and now a researcher at Citadel). He’s also been involved in recruiting in both places. According to him, quant dev requires a mix of programming (in C++ AND Python to maximize opportunities), problem solving skills, and theoretical background (multivariable calculus, linear algebra, statistics, machine learning). The first two will let you break in while the last one will take care of interview curveballs and make sure you can actually do the job. In terms of interview difficulty, he mentioned that CodeForces rating of 1600 -> 70% chance of passing the algorithm bar and CodeForces rating of 2000 -> 95% chance of passing the algorithm bar. Similarly, on the brainteaser side, none of the problems are harder than AIME level. Why quit your job and go back to school? If you take a look at the table of contents for each of the books in the Foundations Of Applied Mathematics series, you'll see that it essentially covers a masters in applied math. Regardless of how good your problem solving skills are, learning that much math will take 3-4x longer if you do it part-time. What if you don’t see it through? This plan is designed to maximize transferable skills in case you drop out early. For example, if you call it quits after phase I, you can walk away with the confidence that you can pass any algorithm interview anywhere; this is invaluable in my opinion. If you leave after phase II, you can quickly build on that mathematical foundation and transition into an ML-adjacent role (e.g. Data Science, ML engineer). Finally, if you tap out after phase III, you have enough of a background to specialize in ML and re-enter industry as a scientist. If this doesn’t sound crazy to you, please join the discord: https://discord.gg/bZh5SND7 We can talk about each phase in a lot more detail.
Do you have a solid understanding of markets? If not, you are effectively trying to become a chess grandmaster by studying math.
I actually used to be a strong chess player (2150 FIDE), but math never helped me make any progress there. Might be misunderstanding the point though.
You are making my point
I don’t think you need anyone to follow you on this journey if you have a plan that can work. With my limited knowledge on how to recruit for quant dev gigs: I think it’s best to simply go to school (either get accepted into a computational mathematics program or get an MBA from a top school OR go get a masters or pHD in mathematics or comp sci from a top program).
Yea…. You can study all you want but you need connections from too schools to land the interview
Please correct your post: Why I am trying to do this? 1. Money 2. Money 3. Money
If I'm being completely honest, I'm more interested in the paid noncompete than hitting some ridiculous unachievable-in-tech comp. Would be able to check off so many bucket list items guilt-free if I got a 1 year paid vacation.
Trust me you would not like it. During non compete you only get the base and no bonus. Bonus is usually 100-200% of the base for quant roles. You would feel shit if you made only one third of what you usually make (as a full time employee vs during non compete). Make sense?
Let's make a telegram for those who want to break onto quant
Down, but I'm not familiar with Telegram at all. If anyone makes one, please share the link - would love to join!
I can assure you we have a lot of non genious quants
Can you explain to me, how on earth do you organise interviews at JPMC? Was invited to a FIRST call thinking it was recruiter screening and accidentally found out it was coding! FYI, that was scheduled for just one day after initial email 😂
Rofl
Since you actually work at one of the big names, what's wrong with how I'm going about this?
Study business, finance, and programming, not math.
I’m very much interested in this! Let’s start a telegram/discord group?
Currently Lead Deep Learning Scientist and have a 1,5 year plan for moving to deepmind/openAI. I appreciate we have different goals but if you’d like to ride together for a while for overlapping tasks feel free to DM.
Whoa, that's super cool! Went ahead and created a discord to discuss further: https://discord.gg/bZh5SND7
i feel really bad for OP, i dont think they realize how many other opportunities exist in life for making money for a fraction of the cost of this opportunity. Moreover, this post demonstrates an enormous misunderstanding of the status games that the quant talent industry preys on. Even deeper, op doesnt even understand the underlying reason why they created this post no real point in even trying to provide advice, theyre too attached to listen
Dudes got a plan Need a whatsapp group or something
Lol I'm down, DM me