Ask HN: Software career at 38 after being a quant
I have been working in finance as a Quant, basically combining some aspects of data science with tool programming. In my free time I worked on side projects including with Flask, Django and Plotly Dash. Basically worked with Python with a pinch of React mixed in. I’m 38 now and seriously considering to move into a software company in the hopes that I can work on probects like my sideprojects. Although I am willing to learn a lot and invest time, I am afraid that I may not be good enough. Should I take a chance and quit my new job as a math and physics teacher (it just does not make me happy)at a high school to persue a position at a software company?
I am at a crossroad now and am happy for any hint if you have ever seen a similar move work out.
Thank you in advance for any reply!
69 comments
[ 5.7 ms ] story [ 128 ms ] threadI’m currently working with a friend who is transitioning from retail investing to software development and the bar for an entry-level role is quite attainable.
Your quant experience could also land you in specialized software dev roles that might have to do with finance or data analysis if that calls your attention.
As for the tech that you mentioned (Python, Flask, Django), roles using those things are alive and well but feel free to explore other languages frameworks to sort out what type of role you might like more (i.e. flask is usually backend while plotly is frontend). Nowadays at least some passing familiarity with JS is useful because of how it everywhere.
Take a look at the threads like “who’s hiring” but see if you can leverage network connections to talk with people instead of sending a hundred copies of your resume everywhere.
Do feel free to reach out (email on profile) if you’d like to chat further and good luck!
When I was working at the acquiring company, they didn’t really have a good place for me. I had more domain knowledge than anyone else at the company but I had less experience with even the basic etiquette of how software reviews work (I.e. GitHub pull requests).
Coding as a quant and software engineering are really quite different. As a quant, you have to optimize for getting stuff done quickly at all costs. In software engineering, there’s more focus on testing, maintainability, and readability.
Eventually I created my own role as a technical product manager and that worked out quite well.
Is it like listening to nirvana when eveyone else is listening to michael jackson?
Regular financial data is stuff like trading volumes and earnings reports. Alternative data is looking at satellite pictures of car parks to see which retailers were busy this quarter.
I would have expected the opposite! I've never been a quant and have mostly done software/data science at smaller companies where there's very little focus on testing/maintainability/readability. I'm sure at bigger tech enterprises those things are stressed more. I figured a hedge fund would also care a lot about those things, at least testing, because a bug can cost a lot of money in a very direct way. How do quant shops get away with looser software practices?
I started as a quant but preferred building software so I left finance to software / data science consulting for a bunch of years and now am co-founder/CTO of a software company that just finished YC (AiSupervision W22).
My advice is to put together a portfolio website with some demos of your projects and well-documented code on GitHub and then create a profile on workatastartup / apply for jobs.
Good luck!
In my opinion, it is a great work very valuable for the society, and that allows to pass your passion for science to children (those who are receptive). It can even be a bridge to some academic career.
However, selling teaching services to the public sector, where price is determined by what a group thinks feels right rather than demonstrated value, can be quite lucrative and on par with the typical software engineering job.
It is true that software engineers have more moonshot opportunities. You will find more software engineers raking in millions of dollars per year than you will teachers. But only a scant few engineers end up in that position. The vast majority of software engineers will land around the same range teachers in the public sector will (with some regional variability, I'm sure).
Where in America can a teacher make 100k+ with <5 YoE?
That just can't be true. No disrespect meant, but I'm not sure if you are under rating what the vast amount of software engineers make or over rating what teachers make. Would you care to clarify with annual take home estimates for both?
Maybe some college professors make what engineers make. That could be true. That is very likely.
Developer salaries are harder to come by, not being officially published in a cohesive form, but levels.fyi suggests a median income of $147,000 (total comp). In my opinion, levels.fyi skews towards big tech, ignoring the little mom and pops that pay peanuts. Other sources suggest the average is closer to $80-90k, depending on who you ask. In the interest of fairness, I'm going to pick the largest number I can find.
Again, not having officially published data, contractual terms are hard to establish, but I think it is reasonable to assume that most developer jobs will expect somewhere around 2000 hours per year – your typical 40 hour work week. In my experience, this is the norm. While I see a growing number of places that allow fewer hours, they come with a commensurate cut in pay. This puts the median developer at $73 per hour.
It is recognized that median and average are not the same, but given the distribution of teacher salaries, median and average should be close to equivalent. For all intents and purposes I think we can consider them to be equivalent in this case. It is not like there is a teacher on the payroll making millions to skew the numbers or something. The pay bands are known.
So, we're in the same ballpark. Within a few dollars of each other. But, as before, I won't be surprised if there is variance across locations. Pay scales can differ significantly from one place to another.
Note this is a based on salary. I would assume delelopers have better total compensation, but there are loan repayment and other things that might affect a teacher.
Here is the US salary guideline :
Teacher salary: $61,820 per year based on an estimated 998,800 jobs reported on taxes.
https://www.bls.gov/ooh/education-training-and-library/high-...
Software Developers : Salary $110,000 based on an estimated 1,847,900 jobs reported on taxes
https://www.bls.gov/ooh/computer-and-information-technology/...
What country are we talking about?
I think that I would be quite good as a teacher, and would enjoy the act of teaching motivated and interested student. However I have enough friends/relatives who are working "capital-T" Teachers to know that actually working in a school is a completely different prospect. Between poor management, ever-changing demands/requirements from politicians/government and terrible parents the actual job of teaching can be a lot less rewarding that people might think.
It's kind of like how some people really love coding, and like playing games so they think "I'll become a games developer and I'll love my job" only to be crushed by the actual process of games development in a sweatshop games company. Or the difference between loving to draw and being a professional designer churning out bullshit for some corporate client.
While you are on contract, you get crappy assignments that no one else wants: substituting for a semester, classes with difficult kids, etc. States differ, but I think this is a common picture for many.
If money is not particularly important teaching at a community college is much more pleasant. Competition is pretty low (because the pay is low), but you get respect and a lot of freedom to teach what and how you like. I know a few folks who semi-retired after a high paying career into a part-time community college teaching and really liked it.
I'm only slightly older than you. You do not want a Jira job. It will burn you out. It's not the same as indie development... you will probably work on tickets, adding features no one really cares about or fixing bugs that you didn't create. It's awful. If you were a real quant, you can probably use your CV to get something a lot better.
It's a valid perspective.
You don't want to end up in a corporate machine working on Jiras after Jiras, which might make you hate your job.
Now I'm depressed :( I get what you're writing, but depending on the project, the people you work with etc., even a "Jira job" can still be fulfilling. Also, as a manager, if you have this "I don't understand how you can do this awful job" attitude towards the developers you work with, I'm not sure if that's really healthy for you either...
I'm always learning new things about various industries. Commercial insurance was particularly fun. All kinds of interesting edge-cases. Didn't feel sleazy. Paid-travel to conventions. I managed to maneuver myself into a secondary role as technical expert and got to work with customers a few times.
All this is engaging to me and outside of work, I'm free to do what I find fulfilling.
A CTO at even a small software company that is writing code is doing their job wrong if they aren't quickly evolving into a role where they don't even touch code.
I've been in multiple management roles at multiple places and, with one exception, the role is much more about managing expectations of leadership and playing political games to make sure your team survives. ICs have some dream that management is about being a technical leader and guiding IC work, but, in practice, this is a very tiny part of the role. The only exception to this I've seen is managers that are grown internally. But if you're coming from outside to lead a team, it's going to be upper management that hires you and upper management that you end up doing most of your work for.
I've known a surprisingly large number of ex-finance people that got "Jira jobs" and they quite enjoy it. Most people leaving finance have very large savings stored so they have a fair bit of extra comfort in knowing that if the job ever starts to really suck, they can just walk out the door. On top of that mindlessly solving problems and getting to check out with only 3-5 hours of work is exactly what many of them want.
hard disagree here. When I worked at Fastmail (less than 50 people), our entire leadership was technical. Even our least technical leadership person (CFO?) was more proficient than most junior developers.
It's a beautiful thing, because business goals get aligned with programming goals, on a long lasting technical project that is very important. No more jostling over what some person with product vision wants versus what needs to happen, refactoring, etc.
I actually think the biggest challenge for the leadership team at fastmail is the leadership part, not the programming part. A bunch of really really smart tenured developers suddenly navigating how to be managers, CEO, CTO, etc, is much more challenging than writing some fixes for a new Apple Calendar quirks. It was cool to see growth from the leadership who took the part of 'ok I am a real C level at a company with a growing headcount how do I do this right the first time?'
I think, having a leadership that still keeps their hands in the dirt is a good thing. Plus they happened to all be the best programmers we had. Leapyears better than anyone else I think.
…and you just identified why writing code in a leadership position is a problem, so actually it wasn’t working. There’s no doubt that leadership with previous experience in the trenches is a benefit, but they should not be actively doing it in that role. Leadership is a full time (or more) position, and coding is a distraction from that.
We have the phrase “can’t see the forest from the trees” for a reason; Leadership needs to be looking at the forest almost all the time, and they can’t be effective if they’re stuck in day to day coding issues (the trees).
What you saw at Fastmail is pretty much standard in the industry: above average intelligence in the worker ranks but way below average ability and intelligence in the management ranks.
No, the opposite. Most of the C level is extremely smart. Smarter than when I worked at a research lab with all PhD's and Professors. All are or have been leads on successful open source projects, contribute to IETF, one guy ran the Perl project for years, and so forth. Very smart crew. One of the best things about working there was being around people smarter than me
Some places allow you to have a say and decision in what comes into the queue and what gets worked on, I think those are a lot better.
If one has to survive in the shitpile we call corporate capitalism, though, then it's better to hold a higher rank, where the jobs are more lucrative, less demanding, and easier to keep.
I've been in multiple management roles at multiple places and, with one exception, the role is much more about managing expectations of leadership and playing political games to make sure your team survives.
Indeed, it sucks. But ICs suffer a high risk of random non-survival. The stress of being a manager comes from knowing what fecal garbage most of the people who run your company are. The stress of being a grunt comes from not knowing. It's hard to say which is worse.
On top of that mindlessly solving problems and getting to check out with only 3-5 hours of work is exactly what many of them want.
I find that, when the work is pointless, there's a hedonic adaptation to not having to do much of it, just as there is with everything else: whatever you have becomes, within 3 months, what you consider yourself entitled to. People who are used to 8 hours per day of pointless work might be happy at first to only have to do 5, but soon they'll grow to hate even that. I, on the other hand, don't mind working if there's a point to it. I want zero pointless work. Unfortunately, the current system, which celebrates the proliferation of malevolent bureaucracies, has nothing for people like me.
why would any company give CTO coding role to someone who has no experience in software.
Good question, but in my experience, it's more the norm than the exception at larger non-tech centric companies. CTO has basically an administrative role there.
If you want greater autonomy, work on hard topics. Not only can you fail more (it's hard after all, right?), but once you deliver a couple times on a hard topic others had trouble with you are given a lot more leeway to improvise. Your only enemy becomes the support requests then..
What I'm trying to say is if one is stuck with a Jira job, they're not motivated enough to convince others they're worth more.
Disclaimer: some jobs give you no say in this and it's near impossible to move out of your little painted corner. Just don't work there.
Do you know the OP? Do you know what they want and like? I doubt it! While this might be true for you, I do enjoy working my "Jira job" at 38.
> If you were a real quant, you can probably use your CV to get something a lot better.
Who makes you the judge of what is better for this person?!
This is my first time hearing that term and I think it's great. Poor Jira though, they get thrown into what is essentially the hell that was created by scrum.
One potential avenue would be to build a SAAS product for Quants with your existing domain knowledge. This would allow you to leverage your existing body of work and do the programming work you want to.
Good luck!
Aim for small companies or early stage startups where they're more impressed by your determination than by your background in the field.
Happy to chat more with you or anyone in a similar situation, email in profile.
1) Make decisions about things, and did the required research to make those decisions (and can talk about what you did and why).
2) Went through the entire deployment lifecycle. Its really easy to just do the interesting stuff then call it a day, but if you have actually gotten to production with something then you are forced to address a bunch of tiny paper cuts along the way which will round your skillset out significantly when trying to get your foot in the door.
Wait so you've already quit finance to be a math/physics teacher?
If you think you will enjoy it, I don't see why not.
>Should I take a chance and quit my new job as a math and physics teacher (it just does not make me happy)at a high school to persue a position at a software company?
I think it's better if you quit your job after you already have found a job as a software developer.
Just as with dating: there’s always someone you are good enough for. And you can build up your career from there.
I hope that this is one way it differs from dating :)
I'd keep on money rolling as a quant, or enter academia, or doing consulting as a quant. Don't waste that knowledge, it's super difficult stuff
My wife&I are your age and have worked at 6 & 5 companies respectively in our careers. Never have we quit a job before having found a new one, gotten the offer, accepted, and gone through background checks.
Why create an exploding timeline for yourself?
Many people do not consider the actual job requirements and skills necessary for the work. It's like, they're just trying to find the best person possible. Some people are even hardcore about this; they test you and judge you quite aggressively. That's cool, and there is some truth to that; there is a bar that needs to be met. But, it's a rather shallow perspective. And, it can have negative consequences. E.g. people burn out, get bored, don't work well with others. There's overlooked value in getting to know a person and positioning them well in a role.
You might be surprised. Your experience from side projects may translate extremely well to a "real" job doing similar things. If you can stay humble, work well with others, and spend time on fundamentals, I think you'll do well in the right environment.
As an aside, within the past year I've sparked an interest and appreciation for biology. In the back of my mind, I sometimes think it would be very fulfilling to work on software/hardware for biotech at some point in the future. Maybe a "second career" type of deal. I have massive "imposter syndrome" about that, but who knows... maybe some day my path will lead me there.
Don’t quit your day job until you have another one though, there’s no point in increasing risk
My suggestion is that you get into data science or bioinformatics. You will be able to pick up the specifics easily, given your quant background, and there is a decent amount of software development involved as well.
While you probably don't need a full year of self-teaching fundamental computer science to be able to do the job, or to interview, you may need to drill and kill algorithms problems on leetcode or similar.