Here at the @DuckbillGroup we're intentional about hiring. We're fully remote, we work largely asynchronously, and we solve fun / fascinating / maddening problems.
We're bootstrapped (no outside investment), so we hire slowly and with great intention / care.
There are two primary issues with hiring junior folks here.
The first: @mike_julian and I are concerned that someone new to the workforce wouldn't have a great experience here. We're not set up to guide folks who are new to office-style jobs. We worry this would set folks up to fail.
The second: "junior" doesn't always mean "new to the workforce," it can also mean folks switching career tracks.
That's likely a lot more tenable here--but that doesn't get around the "mentoring someone intensively" problem.
To be clear: we *want* to hire junior folks, but we want/need to be able to do that responsibly and well or we're setting them up for pain that isn't their fault.
I agree wholeheartedly with this position. As a small ~10 person company, it makes sense. As a publicly traded company it's failing to give back to the ecosystem / train the next generation.
No, and this is probably incredibly unkind to someone somewhere: we're what the AWS Cloud Economics team really should be, but just isn't. AWS TAMs want to join us because they see the Cloud Econ team and wish they were what we are.
Specific to Cloud Economist roles: Solid devops / ops / SRE backgrounds. 2-3 years experience is too junior. "Seen and done a lot" is a prerequisite to understanding our clients' situations.
"I have SEEN some shit" is something all of our CEs will say.
For all roles, we still hire experienced people. We generally look for people with more than five years experience, and prefer 7+.
Honestly? Nothing. Today it would require a fundamental rethinking of how we run our business, and we're just too small for that. We're still early enough that product-market fit is an existential challenge. Working on hiring juniors is a lesser priority.
Because we can teach an SRE how finance works way, way, way more easily than we can teach a finance person how AWS works. @awscloud billing is ultimately about architecture, people, and trade-offs. Those things require engineering experience to see.
Let me give an example of an interview question we used when filling our Principal Cloud Economist role:
"You've got $1m/mo in spend for a large Kubernetes cluster. What do you do to find ways to decrease the cost?"
I will pause here; hit reply and tell me your answers.
Okay, there are basically three tiers of response.
First up, the junior answer: "Buy some RIs or Savings Plans." Sure, okay. You're not wrong,
Second comes someone a bit more in the weeds. They'll start talking about Spot and Fargate. They know what's up in that space.
But the senior folks, the folks we want, come back with "What the FUCK is on that cluster?!"
Because without understanding what's there and how it ties into everything else, you've already lost.
People who approach cost management from a finance perspective will often talk about everything except what's on the cluster, making the fatal mistake of assuming Engineering has their shit together.
(Incidentally, "what the FUCK is on that cluster?!" is basically the @DuckbillGroup core value proposition. It can't be done via software so no VC backed startup will touch it. It requires blending a bunch of skills together into very talented people.)
The kinds of people we're looking for will start talking about assessing workloads and moving them to more cost-effective locations.
Seriously, "Don't assume your client is an idiot" should be engraved on the bathroom mirror of every aspiring consultant walking the earth. Is it terrible? Maybe! But insulting them doesn't get you anywhere.
You don't have a lot of context, and your clients do. Apply your insight to their context. It may be that that $1 million a month cluster is exactly where it should be given their constraints and scale.
Exactly!
"What's that ancient piece of shit doing?"
"Oh, about $700 million of revenue a quarter, so how about you watch your mouth and show some respect?"
You just don't know without asking. This job is way more about people than technology.
And this response (along with the following) demonstrates once again why @patio11 has the role in our ecosystem that he does; this is a founder-level answer.
Something I think gets lost in translation a lot: by default I think a lot of things in tech are wrong. By default I think anything in a client environment is correct.
Both statements are in fact internally consistent.
Exactly. Expensive consultants need to be highly skilled in their specialty. “I don’t know but I’ll find out” is valid and expected, but it can’t be used as a substitute for baseline knowledge in the field.
A teardown analysis of the Duckbill Group's @awscloud bill for April.
Big spenders are RDS (up $200 a month), Fargate, EC2, Glue.
I previously talked about my Lambda Whoopsie that cost ~$80 more than it should have last month. That's a shame badge that's easier to pay than burn AWS credibility asking the Lambda team to fix it.
(I resolved the problem by discovering it was a JS callback / event loop issue so I rewrote the thing in Python. This is a Thought Leader Best Practice.)
The @awscloud marketing page isn't up yet, but the user guide speaks a lot about things that FinServ doesn't care nearly so much about (ETL and data lake issues) as they do other things (supporting insecure FTP from their partners for transaction data runs nightly).
This is a fascinating release, just because it focuses so clearly on a specific industry segment (a vast and lucrative one, to be sure; if you haven't worked in this space you'd be forgiven for underestimating it).
This is very clearly targeted to some customers, not the rest.
So I want to talk a bit tonight about college degrees.
Let me begin with the obvious: I don't have one. Today that's a fun story; my 20s were harder as a result.
It's clear that a degree makes you more employable than no degree.
But I've spoken with a few people lately who aren't happy with their current jobs and are toying with going back to school for a(nother) degree.
Slow down a second, Hasty Pudding; let me unpack that one for a minute.
A degree is expensive; I'm not going to do the math for you on that one. But it's also a lot of time that you're spending not making money, in most cases.
I've spoken to several people with two degrees who are convinced that a third will make them more employable.
Will the powerpoint slides feature a new template?
Will analysts remember that Amazon has a Cloud division?
Stay tuned for another episode of "As The Cloud Yearns"
(Periodic reminder that the single stock I own outside of an index fund is six shares of $AMZN that I've held for years. Not for any hope of financial gain, but because one glorious day I will shitpost via shareholder resolution.)
AWS earnings beat estimates at $13.5B for 2021 Q1 because nobody listens to me and turns their EC2 instances off when they're done with them.