Data Science/Quant/Research roles

Google Hitorigoto
Oct 11 13 Comments

Which jobs are, respectively, the least and most mathematical in work content?

Motivation: I am currently pursuing an MS in mathematics (part-time, top 10-20) and plan to apply for a PhD program in algorithms & discrete math within the next 5 years.



Want to comment? LOG IN or SIGN UP
TOP 13 Comments
  • Goldman Sachs fake engineer
    I can tell you quant on many types of products sees far far more math than data scientist for sure. I'm on the sell side doing soft BS not even trading strategies and I have actually used some nontrivial linear algebra, probability theory, picked up a bit of stochastic calculus, etc.

    Applied ML research can be mathy at the right place.

    All this said, if you like writing code do yourself a favor and just get good at C++ and get rich at a HFT. Industry math is a huge disappointment compared to school.
    Oct 11 3
    • Google Hitorigoto
      BS = Black-Scholes?

      My impression was that sell-side can actually be more mathematical in modeling and optimization, while successful trading strategies often center on some crafty microstructure and execution trick.

      At least stochastic processes (and calculus, differential equations) are grounded in theory, unlike "system design Facebook in 20 minutes" which revolves around guessing the interviewer's subjective preferences.

      Unfortunately, I don't have the specialized hardware and kernel industry experience to get rich quick with modern ultra low-latency. Knowing the metal seems important when you work so close to the metal.
      Oct 11
    • Goldman Sachs fake engineer
      BS = bullsh...
      Oct 12
    • Google natashala
      Any insight that engineer can be quant? Join Bloomberg portfolio/equity teams would help? (Ask for my husband who will join Bloomberg soon)
      Oct 18
  • LG lzml
    Honestly you can't tell by job title.
    Oct 11 2
    • Hulu


      Generally speaking, DS and Applied Scientist sound most mathematically demanding though. Unless Quantitative Analyst is just "less sexy" name for a DS
      Oct 11
    • Spotify Atinlay3
      Yup, in most cases DS == quant, but most of the quants you’ll be seeing are in finance.
      Oct 11
  • New xzYN53
    Hard to tell without context but usually data science is pure bullshit and pays commensurately
    Oct 11 2
    • Google Hitorigoto
      "commensurately" = underpaid, by 10-30% relative to SWE?
      Oct 11
    • New xzYN53
      Often 50%. But a lot of DS are bullshit too so not underpaid
      Oct 11
  • Intel gtfod
    nowdays I feel like there are "data science Quants" who do more data science-y / ML stuff and "traditional Quants" who are more pure stats kind of people. And that's just for buy side. The traditional Quants seem to be more mathematically competent whereas the DS Quants sort of just throw data at ML (guilty) but both types do a bit of each other's jobs too.

    Sell side definitely has more "math" stuff but it's not actually all THAT mathy. Most of the sto-calc is in pricing derivatives and boils down to trying to simplify things to the few stochastic calculus equations anyone actually knows how to solve.
    Oct 18 1
    • Intel gtfod
      Also agree with whoever said it's hard to tell from just titles. Some Applied Science roles are like half data science, half engineering. Some Data Science roles at smaller companies are mostly trivial, whereas they are quite quantitative / math intensive elsewhere. Quant is usually the most consistent (after you figure out which of the 3 buckets the job is for)
      Oct 18
  • Apple / Eng 🐒codemonk
    Probably applied scientist, RS, or RE
    Oct 11 0


    Real time salary information from verified employees