Rethinking the tech writing portfolio in the AI age

Posted on Oct 29, 2026 · 6 min read

A friend asked me the other day to help them overhaul their portfolio. I’m wary of traditional career advice like “Yeah, of course you need samples”, because it feels cliché and useless. So I sat down and thought about what I would put in a portfolio today. I already know that I would take a scrappy blog over a dead-inside collection of impeccable writing samples any time, but that’s not enough for a piece of advice. So let me tell you a bit more about how I think a portfolio should be composed.

In the AI age, a classic portfolio holds less value. Seemingly complex artifacts can be built with a finger snap, and documentation conjured out of thin air. Things are no longer a differentiator between a knowledge worker and a large language model. Nonetheless, I still believe that portfolios, as in curated collections of work samples and, above all, manifestations of your professional self, are still worth the time you invest in producing them. They have to be different these days, though: the good old briefcase full of neatly formatted documentation belongs in the attic. It’s about selling yourself, now. Yes, yawn, I get it, but it’s still true. It always has been. Arm yourself with some courage and be ready to show the world that you’re not an LLM.

Tell stories of how you created the docs that you’re showcasing

Just providing links to samples won’t do, no matter how impressive they read to you. It won’t do because it’s not only uninteresting and hard to interpret for untrained eyes, but also easily reproducible using LLMs (to a certain extent). Instead, tell the stories behind the samples. Your portfolio should contain pieces whose context is interesting and worth explaining. What was the tension behind each decision? Did you provide feedback to the product team? Perhaps you ended up creating a new UI component for the docs to unlock publication? Light the campfire and tell the story of how the doc came to be. You’ll not only show your reasoning skills and taste, but also how you navigated complex conflicts. Focus on quality, not quantity.

Show the kind of problems you like to solve, not just the solutions

If I were on the hiring end of the process, I would pay lots of attention to the kind of problems you like to tackle and the kind of monsters you have vanquished in the past. A parade of docs you have published, while fair and good, doesn’t tell me anything about your skills or how you approach hairy projects; they’re trophies hanging over the fireplace. Instead, talk about the hunt itself. Whoever hires you knows the next job will bring new monsters; what they want to know is whether you’re the kind of person who walks toward them. Don’t be shy about the ones that got away, either: a project that didn’t go as planned says more about you than a dozen shiny pages. And if you notice a pattern, name it, or give yourself permission to create your own job titles, a bit like Winston Wolfe in Pulp Fiction.

Your opinionated take on docs matters more than your grammar

What makes you distinctively human in the Age of Slop is that you are flawed and imperfect. You’re not wired to please every sentient being like an LLM does: over the years you’ve matured opinions, stances, and tastes that reflect your unique journey as an organic writer. There’s never been a better time to blog about your craft, chronicle your explorations, celebrate your successes and analyze your failures. If a blog doesn’t suit your taste, engage in conversations on social media or meetups, or reflect your position on things in your portfolio. If I was hiring someone, I wouldn’t care about how flawless their grammar is; I’d check whether they hold opinions, because that’s a sign of humanity and professionalism.

Don’t let LLMs have the final word on the design of your portfolio

How nice and tempting to let AI create a website for your portfolio. It looks so well done and professional… and also trite. Fonts are the major offender, as most AI-made websites use similar fonts. Leaving the choice of typography to an LLM is a bit like letting it pick your outfit before an interview: don’t do it. Instead, take care of selecting a font (or a font family) that best reflects your personality and core messages. It’s OK if you like the font AI picked for you, as long as it’s your choice to keep it (and a highly ironic one at that). Same goes with color palettes, UX writing (like the copy of buttons and links), layout, and other elements. You’ll find that AI can execute well if you direct it according to your decisions, but that requires that you decide on stuff first. There’s no autopilot. There never was.

Don’t make it about writing, make it about work in the broadest sense

Work is becoming a very fluid thing. The writing was always the cheap part, so the way you want to present yourself to the market is not as a cookie-cutter professional of yore, able to reach 200wpm on a typewriter. No, what you want is to show that your brain is fit for complex knowledge work at the intersection of tech and comms. In a previous post I made the point that technical writing is but one of the many facets of being a writer in tech, either as a docs engineer, UX writer, content strategist, or DevRel. So let your portfolio reflect that range. Show the scripts you wrote to automate a content migration, the linter rules you rolled out, the agentic workflow that freed your colleagues from tedious reviews, and so on.

None of these is writing, but all of them are tech comms. Whoever reads that kind of portfolio won’t see a typist; they’ll see someone who can move between product, engineering, and knowledge design depending on what the problem demands. Someone whose role is a verb rather than a noun. Go out there and show all the verbs in your dictionary.

More posts I’ve written about tech writing and career growth: