Part 1 was me telling the story of how I went from dead-end job to successful DevOps career by accident. Read it if you haven’t already.
Now let’s talk about how, in 2026, to get there starting from scratch.
First off: you’re gonna have to pay your dues one way or the other. There is no such thing as a free ride in life, unless you’re a nepo baby or kid born with generational wealth. For the rest of us, you’ll need to work your ass off.
You’re going to need a roadmap though, and that’s what I’m here to help you with, so you can avoid wasting your time on things that don’t matter, and focus like a laser on the things that move the needle. It still won’t be easy but it will be possible if you don’t quit.
Gut Check
Take a step back and take an honest inventory of your skills, personality and weaknesses: does a high-stakes infrastructure role fit you? Better to answer this directly right now rather than waste time on something you aren’t a good fit for.
Do you enjoy tinkering with technology? Were you the type of kid that would take gadgets apart to see how they work? Did you like building things like Lego, or electronics kits?
Do you work well under pressure? Have you ever been attracted to high-stakes roles like firefighter, EMS, police, ER nurse, etc?
Do you have technical aptitude? Especially in 2026, if you’re young, do you actually own a computer (not just a phone or tablet)? Have you ever installed or run Linux? Do you know what a command line terminal is, and have you used one?
I know the last question seems basic, but I’ve been shocked at the number of Zoomers that are almost completely tech-illiterate, don’t have or have hardly used an actual computer, and basically live their lives on touchscreens.
If you answered no to most or all of those questions, reconsider whether DevOps or any operational role is a good fit for you. Otherwise, continue on!
Two Hop Strategy
The hard truth is this: junior roles barely exist. To the extent that they do exist in established companies, they’re almost always filled by temp consultants. Intern roles only really exist on the SWE (Software Engineering) side, I’ve never seen any kind of operational DevOps/SRE intern role (the exception are CS-pipeline internships, but if you’re going the university path this roadmap doesn’t apply). If you can find one, great, but you probably won’t.
Certifications like AWS Solutions Architect Associate, or CKA (Certified Kubernetes Administrator) have limited usefulness, in my view. The one thing they get you is a pass through possible keyword filters for Hop 1 positions. When I’m hiring mid-levels or higher, I almost completely disregard them in favor of actual industry experience: what you did, the companies you worked at. If getting certified in a structured environment helps you learn, go for it, but be realistic about what it gets you especially given that those exams aren’t free.
The solution is the Two Hop Strategy:
Hop 1: Build relevant real-world experience
Hop 2: Transfer into industry, build career from there
Hop 1: Building Experience
This is the hardest step because it means going from zero to one. It will take the most time and resolve. But it’s the hop with no gatekeeper and no credentials needed, just work.
You have a few options for this one:
Junior helpdesk role, NOC, junior sysadmin, cloud support agent. Anything entry level that is operations-adjacent.
Be careful about generic “IT Support” positions. Sometimes these can be good and useful, but if it involves things like helping 50 year old office workers use Outlook it’s a dead end. If you’re doing real hands-on work at least some of the time with either cloud infrastructure like Azure, or actual hardware, it fits.
Visible, public project that you actually care about. This can be almost anything: a website you’ve built, open source work, an interesting Kickstarter project, or whatever. As long as it’s something real that you’re putting effort into and actually care about (not some dead Github resume-padding thing), that’s what matters.
Focused, demonstrable homelab tinkering. Build something complex with hardware you own. Some random examples: build a full multi-node Kubernetes cluster and run NextCloud on it. Run your own mailserver and email domain. Run a website from your garage. Set up IP cameras and your own self-hosted NVR. There’s no limit, do something neat and impressive that you can talk about in detail.
Couple things I want to note up front:
If you are going for the Junior helpdesk route, be aware that your goal IS NOT to just do the work, get “Meets Expectations” every performance review, and somehow expect the next hop to just happen. Your goal is to use that first job as an EXCUSE and OPPORTUNITY to do the real work that matters, which means getting the official work done early and then use the remaining time to do actual infrastructure work and build up your experience. This could be as simple as using the company’s AWS account to provision test infrastructure and get familiar with networking and security concepts, or whatever. Point is: do adequate work (don’t get fired), then put the bulk of your time and energy into educating and bootcamping yourself on company time. If you play this right, you can come off to your boss like an ambitious guy who’s interested in the work and going above and beyond. Just be careful not to run up the AWS bill or do anything stupid, security or ethics-wise.
If you’re going the homelab route, you need to be disciplined and carve out time. This is why the helpdesk role is nice, because you have a luxurious 8 hours per day to get the boring work out of the way so you can focus on the things you care about. If you’re working some non-tech job, that means you have to carve out additional time to get this done. Not impossible, but will take longer.
Timeline is 1-2 years for the helpdesk training-on-the-job option. 2-3 years for the homelab option if you’re dedicated about it. Both are easier the younger you are with fewer responsibilities: kids and family make it harder and slower. Fast learners have an advantage: the main requirement is skill acquisition, so judge by that and not strictly calendar time.
A Word on AI Tools
AI is amazing and it can do a lot. Too much. You might be tempted to build your homelab by buying an AI subscription and then letting an agent do it over the course of a few hours. This misses the entire point of the exercise. You are building systems as a proxy and substitute for a formal certification program: the point is the learning, not the end state. If AI does it all, you’ve accomplished nothing and you will fall apart at the most basic intro questions of any interview.
AI can be super-helpful at getting you unstuck when you run into tough problems. I wish I had it back in the day when I was mystified by Linux kernel oops messages or couldn’t figure out how to script something. Just be sure that each time you use AI, you spend the effort to learn WHY the problem occurred, how to prevent it next time, and WHY the AI-suggested fix works.
Foundational Skills
Your goal in Hop 1 is to get enough skills and experience that you have something to put into a resume, and hopefully make it interesting and unusual enough to convince a hiring manager to take a chance on you. Beyond that, you want enough of the basic skills down so that you can pass a set of interviews and actually do OK at the job initially, while you ramp up.
Here’s the bare minimum to consider Hop 1 complete:
Linux sysadmin experience. I don’t care if you’ve only worked in Microsoft shops at a helpdesk, there’s no excuse not to get some Linux experience. “Sysadmin” means the ability to set up a new system, set up users and permissions, install and reconfigure software, add or remove hardware, reformat disks, and fix the system when it breaks.
You should know where system logs are and how to read them. You should know where to look when things go wrong and the different types of logs (dmesg, journald, application logs).
Unix CLI. You need to be familiar enough with the CLI to do sysadmin tasks. You should know what
ps auxordu -hs |sort -hdoes offhand and be able to explain it. Knowledge of how to use pipes, standard utilities like cut, awk, uniq is essential.Bash scripting. Building on the point above. You don’t have to be a programmer, but you should have some ability to write basic bash scripts to automate tasks. You should understand shell functions, variable expansion and how to pass parameters from the command line through to the logic in the script.
Hop 2: Move into industry
This is where your prep work in Hop 1 really matters. If you don’t pay your dues adequately, you will run into a brick wall here. Better to spend the time in Hop 1 to do it right rather than prematurely try to move into Hop 2 and fail completely. One clue: if you’re anxious to get on with Hop 2, it might be because you’re not actually passionate about technology or fixing things and this is all just an income/job play for you. Nothing wrong with maximizing income of course, but if you’re not a die-hard tech nerd this probably isn’t the right path for you. Figure out sales or e-com or something.
Remember when I wrote that there are basically no junior positions? That’s true in established companies, but your goal is to find the smaller companies, startups, non-tech companies that have some kind of tech department, and so on, that would be willing to take a chance on you.
And let’s be real: you have zero chance of getting into Meta, Google, Apple, etc at this level. Maybe a tiny chance at Amazon if you land on a manager who wants to hire-to-fire, but you don’t want that. You’re looking for the tiny company, cash-strapped startup, or one-man operation who wants someone cheap who can work hard, has initiative and will grow into the role.
Here are places to look:
Hacker News “Who’s Hiring?” threads. They post this monthly. Scour the posts looking for infrastructure/cloud engineer/SRE/DevOps roles.
Social Media. X is probably your biggest opportunity area. Facebook is mostly boomers but worth looking at. You could try Bluesky but the vibe is…different, so be aware of that.
Niche job boards. There’s a surprising number of these out there for specific industry areas or technologies. For example: bitcoinerjobs.com, golangprojects.com. Search for “<technology/area> job boards” and see what comes up.
Standard job boards. Don’t neglect these, even in 2026. Indeed and ZipRecruiter can sometimes have interesting posts.
Relocation
In this Hop 2 job search, you’re going to need to give yourself every possible advantage. Even if you succeed in everything up to this point, if you put too many conditionals on the job you’re willing to take (“must make $X/year”, “must involve technology Y”, “needs to be fully remote”) you can shoot yourself in the foot. When you find that hiring manager that likes you and is willing to take a chance, you need to jump on it immediately.
Commit up front, before even starting, to make this happen. If you don’t live in a top 5 metro area, your chances of finding a position that will both accept your junior level, and also allow you to work fully remotely, is way less. You likely will have to pass on what otherwise would be golden ticket opportunities. Even if you DO live near a major city (say, Chicago), if an opportunity in New York opens up, you don’t want to have to say no to it.
Keep in mind that all metro areas are not the same for your purposes. The areas with strongest tech and startup cultures are San Francisco, NYC, LA, Austin, Seattle. In contrast, major cities like Houston, Atlanta, Phoenix, Miami, not so much.
Personally, I’m sympathetic and conflicted because I’m a fan of strong, local communities and I hate that young people often have to leave their families to follow their careers. That said, we don’t live in a perfect world and this roadmap is all about how pragmatically to get where you want to go. So be realistic.
In the Job
So now you’ve gotten your first industry job. Congrats! Now get ready to work harder than you likely have ever worked in your life. Especially if you’re in a startup, expect to work stretches of 10-12 hour days, 5-6 days per week plus on-call duties. You’re likely making just enough to survive, not much more. If this sounds like hell, keep in mind that this job is your paid bootcamp. Unlike a university or private bootcamp that you have to pay for, this bootcamp pays you. Your goal is to grow in this position for 1-2 years, then make your step-change transition into a mid-level role where you will greatly increase your income and career.
Patience and willingness to learn is key. Try to keep a sense of gratitude and humility: you don’t really know what you’re doing, you’re going to screw up, and these people took a chance on you. Own your mistakes, don’t be arrogant or entitled, and do everything you can to both level up yourself and also contribute to the company to make it a better place and succeed.
How do you know when you’re ready to move on? Probably when you start to feel comfortable, and when start to feel limited both in terms of work assignments and advancement. You’ll likely only get incremental promotions and comp increases in this role, so changing companies into a mid-level role will give you way higher upside.
Polish up your resume now that you have real, demonstrable industry experience and see what’s out there. Hit LinkedIn, find industry recruiters (CyberCoders, etc) and apply directly to companies. It will still be an uphill battle for the highest tier of companies like Google and Meta, but it’s not impossible either depending upon your skill.
If you play this right, you can likely double your income. Maybe more, if you include equity and perks. Tech roles like this are some of the only W2 roles in existence where you can routinely get meaningful equity (BowTiedBull makes the valid point that this is increasingly the only way to generate real wealth). This is the step-change payoff that you’ve been working for years doing the Two Hop method. This is also why the operations area is relatively uncrowded: paying dues isn’t fun and not a lot of people have the drive or persistence to make it work. If you’ve done it, be proud.
What’s Next
Next up: the homelab roadmap. What to build, in what order, and what skills each project proves. This is the 18-month version of the one that took me a decade to stumble through by accident. Subscribe and you'll get it the day it drops.
I'm opening paid mock interviews soon: a full SRE/DevOps interview loop run by me—the guy who actually runs these loops for a living—followed by a written scorecard telling you exactly how you did, where you can improve, and my honest assessment of whether you’re ready. Slots will be limited and waitlist gets first crack. Email hello@bowtiedcessna.com with "coaching" in the subject line to get on it.