working as a machine studying engineer at a Big Tech firm.
On paper, I had a dream job:
- Versatile working
- Good and pleasant colleagues
- Nice perks and advantages
- Good work-life steadiness
- Barely any conferences
- And my compensation was nicely over $100k
Regardless of all of this, I at all times felt one thing was lacking.
I initially thought it was a section and I wanted to present it extra time, however the feeling by no means appeared to go away as months handed.
If something, it grew stronger, and I began to really feel unmotivated.
I like this area a lot; I’ve actually been running a blog and filming YouTube movies about information science and machine studying for over 3 years, however this previous yr, I haven’t felt the identical enjoyment.
This actually bugged me, as I’m nonetheless comparatively early in my journey, and there are such a lot of issues left for me to be taught.
I knew one thing needed to change.
I wished to get that keenness and pleasure I had solely two years in the past.
So, on this article, I wish to go over why I finally stop my machine studying engineer job, and provide an alternate view to what these “dream” jobs are literally like.
For sure, that is my opinion solely on my brief expertise in a single crew and shouldn’t be taken as a mirrored image of the corporate or its folks.
Tempo
Regardless that Large Tech are clearly know-how firms, that doesn’t imply they transfer that quick relating to testing and iterating concepts.
As firms develop, they naturally rent extra workers and add extra ranges of their company construction. Subsequently, forms slowly creeps in.
There may be not a lot you are able to do to keep away from it.
This occurs when the corporate is often doing very nicely and making vital income.
Because the outdated adage goes:
If it’s not broke, don’t repair it
Due to this fact, these firms change into much less more likely to check new concepts or methods to guard their backside line.
They’re much less prepared to expand, riskier swings, so to talk.
I get it, it makes whole sense.
Nonetheless, for people like myself, this sort of tradition merely doesn’t swimsuit me.
Fact be instructed, I’m a really scrappy, pragmatic and action-oriented individual.
I don’t trouble testing each single intricate element, or spending an excessive amount of time on utterly random “what-if” questions and happening the analysis-paralysis rabbit gap.
The very best technique, in my view, is to have 80% confidence in your concept that it’s going to work by way of offline testing, worst-case state of affairs modelling, and so on., after which ship it into manufacturing to see what occurs.
Some folks might imagine that’s reckless and considerably silly.
That’s nice, I’ve realized you’ll be able to by no means fulfill everybody.
To me, this strategy is far more enjoyable and motivating as you get to incessantly see your creation exit into the world.
Positive, typically you’ll utterly strike out, however that’s the purpose of this course of.
It’s iterative, and also you be taught and construct a greater product subsequent time.
Sadly, this manner of working doesn’t align with the tradition of enormous firms, or no less than not with sure groups, from my expertise.
Put bluntly, it didn’t align with how I labored, so I struggled to remain motivated.
Lack of Goal
It’s a cliche to say you’re only a small cog in a giant machine, however that’s precisely how I felt.
Just a few months in, I realised that my work didn’t actually matter all that a lot.
Positive, it generated impression, however within the grand scheme of issues, it was only a drop within the ocean.
Whether or not I used to be there or not, the corporate would run easily, flip a revenue and maintain cranking out income for shareholders.
Don’t get me flawed, I perceive that it’s a excellent instance of excellent enterprise and the way an organization needs to be run.
Nonetheless, it made me really feel somewhat nugatory and missing objective. Something I did was principally futile, and that basically hit my motivation.
That is in all probability coming from some ego, however I wished to really feel actually valued and finally accountable for the place the corporate goes.
If I go away an organization, I need them to really feel it.
Being helpful is what brings me objective, and I finally didn’t really feel that all through the previous yr.
Inner Tooling
This can be a slight rogue one, however many of those giant firms have a great deal of inside tooling that they’ve developed through the years to spice up productiveness.
For instance, as a substitute of working with AWS immediately, the corporate has its infrastructure engineers construct wrappers round AWS to make its core companies simpler to make use of and to higher handle position permissions.
Google is one firm that’s infamous for having many inside instruments, however many sources state that they are very good.
Whereas this sounds nice on paper, you don’t discover ways to truly use issues like AWS correctly, so that you don’t choose up transferable expertise that you may apply in different roles when you resolve to depart.
In my expertise, there have been many inside instruments for elementary expertise I wished to be taught:
- Utilizing cloud methods
- Constructing mannequin deployment infrastructure
- Organising automations on Git/GitHub
These had been simply given to you on a plate, and I didn’t need to assume twice about it.
Positive, it does enhance productiveness, I gives you that.
However I’m somebody who needs to essentially perceive what’s going on underneath the hood on a regular basis, as a result of when one thing breaks, I wish to know the best way to repair it.
I didn’t really feel I realized a lot from this, and that’s not what I need at this level in my profession.
Small Scope
There have been round 100 machine studying engineers throughout the corporate, and round 5 instances that quantity throughout the entire information, machine studying and science organisation.
Given this variety of workers, most of the merchandise and algorithms had been very mature, to the purpose that it was extraordinarily tough to squeeze out any additional features or make a considerable impression.
It isn’t essentially a nasty factor, and it’s clearly my job to seek out methods to enhance.
It’s what I used to be paid to do.
Nonetheless, when you have got a whole lot of individuals working or who’ve labored on the identical algorithm for over a decade, the scope of the enhancements you can also make may be very small.
The one actual various is redefine the best way to strategy the issue. However, as I stated firstly, no established, worthwhile firm goes to wish to spend a yr redesigning a whole system.
It’s merely not sensible, neither is it value it in senior management’s eyes.
Numerous the work I used to be doing was extra upkeep and conserving the operation working.
There wasn’t a lot scope to implement new options or algorithms, and over time, the work grew to become stale and unmotivating, as I discussed to start with.
What’s Subsequent?
The simple route was for me to remain, finally earn a promotion to senior machine studying engineer, and have a snug, well-paying job for the following decade.
However the place is the enjoyable in that?
I’m solely 26, and if there was one factor I’ve realized about myself up to now yr, it’s that I don’t shrink back from dangers and I’m far more entrepreneurial than I initially thought.
I wish to construct huge issues that nobody else has, and make my little dent on the planet.
Many individuals will roll their eyes or scoff at me after I say that, which they’ve executed earlier than proper in entrance of me.
However that’s the worth you pay when you’re delusionally optimistic and need issues that others are too scared to attempt and even say.
So, I’ve determined to do a full 180. I’m going from Large Tech to being the sixth rent at a startup.
Large change, with huge danger. However because the saying goes:
Nothing adjustments, if nothing adjustments.
I’m very excited for this new journey, and may’t wait to assist construct a unicorn.
One other Factor!
Be part of my free e-newsletter the place I share weekly ideas, insights, and recommendation from my expertise as a practising information scientist and machine studying engineer. Plus, as a subscriber, you’ll get my FREE Resume Template!
Dishing The Data
Weekly emails helping you land your first job in data science or machine learningnewsletter.egorhowell.com
