Five years as the sole UX guy at a cybersecurity startup, shipped 30+ features(internal and external facing), a Custom Dashboard that got a client who genuinely disliked our product to say "this is wonderful," and an LMS training course I built twice over that certified 160+ people.
IT teams in cybersecurity are perpetually underwater, getting them to sit down for a customer interview was its own project. And knowing tools like the Adobe suite only got me so far, because I was the only one on the team who knew them. A tool nobody else can use isn't a very effective tool.
A team is only as strong as the tools everyone actually shares. The moment tools fragment, the work fragments, and bottlenecks start to show. I also learned that startups need to hire slowly and deliberately. Too many new voices too fast just rocks the boat and taxes everyone with constant redirection. And on sales calls, sell what the platform already does well, not what it might do someday.
Pay closer attention to human nature, not just pain points. A product can solve a real problem and still get rejected if it makes the customer look bad by surfacing the gaps they missed, they'll walk away from it. I'd also move faster to identify actual decision-makers and rule out prospects who were never going to pay, instead of spending time on people who couldn't say yes.
Took a proof-of-concept that had stumped the team for two years and shipped it with a 2-part architecture instead of the originally proposed 3. Fewer moving pieces meant we solved it in a few months, then spent the rest of the year porting it across platforms, Windows and RedHat both.
I came in strong on architecture but weak on the actual mechanics of the web how requests get served, what happens at a given port, the full authentication and authorization handshake. By the end, I still couldn't touch the frontend. I could design the system but not dress it.
A beautifully engineered backend means nothing if the frontend looks like an afterthought and ours did. That gap is what eventually pushed me toward design. I also came to respect C and lower-level tools in a way I hadn't before; there's a clarity to working close to the metal that bloated frameworks paper over.
I was juggling a dozen open terminals just to keep different servers running. I wish I'd invested early in tmux and nvim, it would've saved real time and made the whole workflow feel less like managing chaos.