Post-startup hedge fund hiring remote: mature hackers/makers/modder types


New Member
Dec 2, 2019

I have been lurking on ServeTheHome for a couple of years now, but just did not have time to participate much until now.

I would like to introduce myself and the company I work with and then describe some of the opportunities available and the sorts of people we are looking to hire. Bottom line upfront - we are hiring remotely across a gamut of roles and can consider people in most locations. Full-time is ideal, but for very capable people we can be flexible - part-time or longer-term fixed-term contract (less than a few months is not very efficient for us).

Symmetry Investments is a post-startup alternative asset manager - c $8.9bn of assets under management, with around 230 people. Founded in 2013 (managing money since 2014), we have offices in London, HK, Singapore, and Jersey, with a US office under way.

I am responsible for technology across the firm, and co-direct research. My path to the role was unconventional, and our approach to technology is informed by the different path I took; my first contribution to public-domain software (open-source didn´t have that name then) was to Tom Jennings´ Fidonet BBS in 1989. I used to run a BBS of my own - Absolute Zero - and edited the "security-research" section on 01 for Amiga, a popular London bulletin board. You can read some of my more recent writings on Quora, and I have a Twitter account - @Laeeth.

Along with the obvious benefits of technological progress since the 8-bit days, I feel that we lost something. The BBC Micro at school came with a pretty good user-guide, including a schematic. You could hope to understand every part of the machine and with a basic knowledge of electronics and the technical data sheets on the chips you could even design and build your own computers (now we would say embedded devices). Today everything is so complex and there is layer upon layer of abstraction and that mindset of figuring things out isn´t generally how we approach hardware today, especially in the enterprise. Have a problem? Well in that case you surely need a solution, and since you are a serious enterprise then you must pick from one of the plethora of predigested enterprise solutions available for that purpose. You must be crazy to want to understand it or do it yourself - leave that to the experts.

Our experience has been that although a basic approach of favouring getting your hands dirty may not seem optimal, might even seem wasteful from a narrowly commercial perspective, that trends in enterprise technology have gone so far that the opposite is true. If you understand something then you can reason about it, change it, know what is easier than how it sounds and what is much harder. Or to put it into enterprise-speak (I studied economics), such an organisational capability creates adaptiveness and strategic optionality. Practical experience is also the source of innovation.

When I became more directly involved at Symmetry in 2017 we used perforce. I think Google use perforce, but they have a lot of tooling on top and we did not. So nobody read anybody else´s code - nobody even had access to it. I set up a Gitlab server myself because at the time we lacked capabilities (to work with git, with Gitlab, and even to build and manage Linux servers). That isn´t typically how somebody in the role I had at the time (one of the three people running Symmetry) would approach things but it was by far the best choice given the constraints.

That´s changed a bit, especially on the development side. We wrote our own domain specific language for reporting and integrating internal systems. We exploit the metaprogramming and compile time function execution capabilities of the D programming language to automatically wrap and register straight code written in D. So for the maths part of the standard library we loop through all the symbols in std.math from Phobos and register them. D has some libraries, but C++ has more. So we made an internship project to call all of C++ from our language using CERN´s cling and it basically just works - talk before the Compiler Research Group at Stanford IRIS-HEP is here, and Alexandru wrote a piece for the LLVM blog that came out just recently.

We have Andrei Alexandrescu working with us on the development side (he joined last year) and recently Amaury Sechet, author of the SDC experimental compiler, co-creator of Bitcoin Cash and technical lead of Bitcoin ABC started working with us. We are quite involved in the D language community and have open-sourced (and develop openly) quite a few internal projects.

We started hiring remotely from the moment I became involved, and the pandemic accelerated that - we have been forced-remote for now since Feb 2020. I would say that the development side has been remote-first for a while; and the rest of the company (trading, legal, compliance, accounting, operations, research, and HR) is moving towards remote-first. The employee handbook - like many other internal documentation projects - is now markdown in Gitlab. No big deal for a tech company, but this really is not the norm in institutional finance and we overcame many challenges to get here.

Where we have a gap is for what we used to call infrastructure, a term which covers a gamut of roles. Suppose you have a rogue DHCP server somewhere in the office. We don´t control the switch or have access to it. Quick - how would you track down where that is? It´s of course not difficult to figure out how to use wireshark or tcpdump, but figuring things out like that is a very different mindset from the bureaucratic enterprise finance technology mindset that seems to dominate in people we might hire from investment banks.

Or suppose we want to research server builds. What are our choices for 1U dual EPYC 7002 series with U2 flash? Or we would like 10GBaseT switches you can run Cumulus Linux on - what are the possibilities? Or we have people that know Linux, but nobody knows Cumulus yet. We would like to move to netops, and yet on the ops side of devops we are just beginning to learn.

[Some background is that we outsourced our desktop support, server support, and provision of managed servers to an industry-specialist managed service provider that we have now outgrown. It takes time to build new capabilities to take back control, and we are at the early stages of that].

I would say the distinct components of roles that might be of interest to readers of the forum and their friends are:
- frontline desktop support (mostly Windows, but share of less-technical Linux desktop users is growing)
- problem-solving for desktops - hmmm why is the latest Windows upgrade wiping hard drives including logs on some machines; why is Graham´s computer so slow
- front-line server support (we use SaltStack but our understanding and use of it is nascent)
- devops roles - working with very good developers to monitor + administer servers using config + code
- problem-solving for servers + network
- research for desktops, servers, network and physical office (how do we actually make our offices safe for example)
- administration: managing equipment registry; sending out hardware to people, etc
- any other problems that come up that need to be done that someone can uniquely do better than others
- client relationship managers (some people are very good at their job, and not very good with computers and require a higher-touch approach; or they may just have very little time)
- documentation and business analysis
- organisational design, management of all of the above

Many firms today create a box describing a role and look for people who can fit the box. You had better be able to do everything it says on the box -- and don´t you dare step outside of it. That is not how we have chosen to do things; we do not have hard and fast divisions between roles and we allow people to get varied experience although we do make _distinctions_.

Who would be my ideal sort of person? Well, everyone is different. But I would hire Rachel By the Bay in a flash - a serious hacker that likes helping people.

Credentials are interesting pieces of information, but they are only that. One chap I hired - a contributor to the D programming language - did not go to university; he did not even finish high school. Another was a baker in Spain. He studied technology, found the job market in 2008 difficult and so worked for his father´s bakery business until recently. But he was also heavily involved in the Linux Kernel Newbies project (documentation) and contributed some patches to the kernel itself. He would not have gotten past HR in many other financial firms, but I snapped him up - and he is doing an outstanding job on technical documentation with us. Mature eighteen year old that got into working with hardware to earn extra money by buying, fixing, and reselling computers when he was at school turned out to be a very good hire too.

This being said, I think we have pretty high standards - we just care about different things from some others: you should be able to figure difficult problems out and generally, you know, like and have a flair for working with computers: *mechanical sympathy*. We assess this as part of our selection process, but if you can point to work you have done that is public then it is easier to assess and make sure not to miss outstanding talent.

I said I was responsible for technology across the firm. I will be co-CEO shortly, so I have quite a lot else to do. But last year I set bonuses for the people we have in these roles, and funnily enough I bumped them up quite a lot higher than had been previously proposed. Why? Because although we were lucky enough to recognise the pandemic as the challenge it was as early as January 2020 there was an incredible amount of work to do to be ready for it in time, and without the hard work and creative problem-solving of people in these roles, maybe we would not have had a company at all. So I think there is a recognition of the importance of technology generally, and these roles specifically.

We pay pretty fairly initially and over time recognize accomplishment by paying well to very well. Our mindset is relational rather than transactional - we do not have a mindset of narrow short-term optimization. In the past years I have found that approaching a goal indirectly - by moving towards what you are drawn to - can be much more profitable than short-time calculation and optimization.

The longer-term stakes are high, but we try to keep the short-term pressure as low as possible, although we might not on every occasion succeed, particularly when it comes to supporting a trader who has a large amount of risk and is being let down by a failing computer. Even in those conditions, most people are pretty reasonable here - you don´t get much shouting or general interpersonal obnoxiousness. Many roles we are filling now are not like the supporting stressed-trader role though - researching and building new choices does not have a short-term time-element aspect at all.

Let me know any questions. Email me lisharc at symmetryinvestments dot com and cc jthompson at symmetryinvestments dot com if you would like to apply. As well as a bio, please send links to publicly-accessible work you have done (code, designs, blogging, social-media etc) as well as a note describing why you are interested in working with us, and what you think you could bring.


  • Like
Reactions: marcoi


Well-Known Member
Jul 21, 2017
I'm too busy with what I am doing right now to even consider working for someone else, but as a business owner myself, I really like your approach to IT as well as your chosen talent. I'm one of those people who don't have the paper credentials, but if you look at what I have done in my life and the challenges and solutions I've been able to come up with, well, I even impress myself sometimes. :D

Best of luck in your search and maybe in a year or two I'll reach out and see what type of fit we can be. :) And best wishes for the future!
  • Like
Reactions: Laeeth


New Member
Dec 2, 2019
Hi Samir.

Thank you for the kind words. My experience has been that character and resourcefulness are built through the repeated experience of overcoming adversity, so perhaps there might be a link between having encountered challenges and having built up the ability to solve problems.

Please do get in touch if you are ready at some later time. We will be hiring for a while yet and for capable people can accommodate particular circumstances if we can think of a way to do so whilst being commercial.



Well-Known Member
May 23, 2015
You know @Laeeth ... I get the feeling that you're looking for people who are pretty much...jack of all trades, when it comes to technology. That's easier said than done.

The number of people who build software AND understand how a computer is actually put together (and understand the innards) is very small.
The number of people who have the above skills and even know what port flood is when it comes to networking, is even smaller.
The number of people who have the above skills and even know how to produce a golden image for get the idea. And the moment we venture into cyber security...

My point is, "IT" in general, has become complex. Very complex. There's a reason you're looking for such talent... :) If it was easy to find, you'd have found them by now.

And're on Wall St (figuratively). That's either a lucrative draw or a complete turn off for a lot of people. The kind of talent you're looking for will be highly skilled, highly opinionated, highly vocal and usually won't take BS in any shape or form. Money is not even the primary driver for them. They do what they do for the passion of it, money is a by product.

Let's make it real. Would you pay a trader $1M/yr (chump change in your/mine world) or the person managing your data center?


New Member
Dec 2, 2019
Just briefly, as it's late here.

I wouldn't say necessarily Jack of all Trades. But I took Electronics A Level although I never studied for it - I used to hang out at the local radio and TV repair shop and then I taught myself the rest. And I think knowing the principles of a domain can be very powerful because it gives you the ability to figure things out without necessarily being taught them. It's true that the incidence of people who think like that is lower amongst professionals these days. But you don't necessarily need to be a professional currently to be able to learn how to do it. And more than anything else it's a big wide world and we don't need to hire many people and the internet makes it easy for highly unusual people to find each other. And my approach has worked for other roles and I think it's harder to find pure programmers than the roles I describe above. Maybe I am unrealistic - I am willing to take the risk of trying. I don't think so though - plenty of people who spend time on forums like this one could do valuable things for us.

The reason I am looking for such talent now is we grew very quickly on the business side and I started with programmers and now have a bit of breathing space to work on other roles. I didn't get involved in hiring for this area before really. Honestly at the moment programmers are the easiest role in the firm to fill because I bootstrapped by hiring people myself early on and now we have very good people and capable people want to work with capable others and there just aren't that many places like that these days. Everything is also hyper optimised lest people make mistakes and for us we think about errors in terms of the economic downside (usually less than people fear) and the upside from creating an environment where learning is possible. In an age where people lack courage, a little bit of daring can go a long way.

I mean our internship project was to call most of C++ from our little internal functional DSL. Something I hacked on and Alexandru finished. There aren't that many places that do that sort of thing.

I don't think we are looking for some kind of full stack programmer who is an expert on multiple CPU architectures, hacks on open source RISC design for fun, has built their own computers, solved decades old bugs in Excel that Microsoft couldn't fix etc. The Chuck Norris of support!

No - we are just looking for highly capable people that like working with computers, solving problems (and for some roles helping people), are okay with having to learn quickly, are intrinsically motivated and would like to work with us.

Well in London we were in the old house of crazy old Isaac Newton. That's quite a long way from Wall Street in so many ways. But we started preparing for covid in January and the office has been closed since February 2020. I read Reddit and I can recognise patterns and that was enough to see what we needed to start doing. The looks I got from colleagues when I said we must start to prepare for everyone to work from home for a year! Btw here's a paper I wrote with friends on vitamin D. It's an evolution of a paper we sent to certain people in May.

I think maybe I didn't get across the sort of people we hire well. We are looking for intrinsically motivated people and it's something we even try to test for with our assessment test. Make your own mind up about me, but you might have certain preconceptions about what everyone in a certain industry might be like and maybe I would not argue with you about the average but it's quite a big world and so there are creatures one might never have imagined, very different from what you might think.

Opinionated, vocal, not taking nonsense. I am not sure that's something new to us in the way you think it is. The few more enterprise guys we had were horrified when my first hire - his first job but he is a very good programmer - argued with me in front of everyone. If people won't disagree with me, how can I know where I am wrong. It's way better to know before you do something stupid than after. And even if it's after then the sooner you realise and correct your mistake the better.

Would I pay the guy that manages our data centre a million dollars a year? Well right now that's still sort of me - it's not used for production yet so don't worry - so it's a difficult question to answer. But I built the machines (with some help later on), configured the network, figured out what container technology we wanted to use (we use LXD. Cool thing we just did is run libvirt inside LXD so you have one set of images to manage and even Windows VMs are just containers). Wanted to be able to manage zfs from our little language (not so little by now) so I wrapped libzfs_core for D, wanted to manage LXD so I wrote the first version of wrappers for that. I didn't rack the machines though - I have a bad back.

I don't need to get as involved these days on the software side because we have very good people and I also don't have much time. But early on bootstrapping by doing things that don't scale was necessary because we just didn't have the right capabilities internally.

I really didn't have many choices early on. But a benefit of getting your hands dirty is that you develop an understanding of what the words refer to, so you don't end up in these awful situations where what you say sounds good but actually you have no idea what you are talking about because you don't have any mechanical sympathy for the system you are discussing because you haven't worked on it yourself!

There's also a benefit of a degree of conceptual coherence in the emergent design of a system. It's not like Frankenstein's monster - behold, it moves!

I suppose that wasn't really what you were asking though.

For the time being we are in a free market economy and we are a commercial endeavour within that context, so I suppose that right now it's true that traders have higher compensation on average than technology people. We manage money for institutional investors like pension funds and insurance companies. Undertaking an interesting social experiment of switching compensation scales wouldn't really be in their interests or I suppose in ours because we would lose trading talent in time and our returns would suffer. I've done both roles, by the way, and I am not sure it's expectationally better to be a trader because most traders don't really make it. You just never hear their stories.

Looking at it the other way though, we are a commercial endeavour and it's an advantage to be able to recognise talent and pay it what it is worth, which should be more here if we can organise ourselves in a way where talented people are able to be most effective. If people can accomplish things we can recognise that in pecuniary terms very well over time. We aren't driven by credentials or pay grades and don't have these rigid boxes that we put people into. We try to design jobs considering the work we need done and the interests and capabilities of people we hire - not that there aren't less appealing bits to any job, of course. Maybe sometimes we even tolerably succeed at that.

Would I pay the guy who manages our data centre a million dollars? Not right now because we don't do HFT or anything and in fact most of that right now is build servers. It's a part time job and not rocket science. We wouldn't be worthy of someone in that bracket of capabilities because we just don't need that much at this stage. Maybe we will some day though, and if that's right we would pay well for that job because it's an important job! We certainly recognised the contribution of people working in support last year.

The status of technical people in our firm is a bit different from peoples image of finance. It's not a case of Eloi versus Morlocks. Supposing it were - well there's a Morlock helping run the firm.