Hiveon’s Lead Frontend Engineer Anton An
According to consulting firm Protiviti, 70% of IT leaders believe technical debt undermines their company’s ability to innovate. Over a third of IT budgets are spent dealing with its consequences. In product development, this problem becomes especially evident over time.
Anton An, Lead Frontend Engineer at tech company Hiveon, has seen firsthand how technical debt manifests on the frontend — and knows what can be done to keep it in check.
While many non-engineers think of technical debt as a backend issue — with slow queries or inefficient databases — Anton argues that the frontend is no less vulnerable. “People often think the frontend is just buttons and forms — what could possibly break? But UI has its own architecture. And mistakes there can be just as painful.”
He compares it to cutting corners during home renovation: “It’s like deciding not to level the walls properly, just to finish faster. Seems harmless — until the wallpaper starts bubbling. That’s exactly how frontend tech debt works. You rush a feature, add an outdated plugin, hardcode business logic into a button, or write ‘temporary’ code that turns into a mess a month later. Then a small change somewhere triggers bugs in completely unrelated parts of the app.”
Anton points to three clear symptoms of accumulated technical debt: slow development speed (“what used to take a day now takes a week”), a growing number of bugs (especially regressions), and team burnout. “Developers become afraid to touch the code. They feel like they’re doing blind surgery every time.”
The most dangerous form of technical debt, in Anton’s view, is chaotic state management. “When a project uses Redux, Context, local state, and an experimental library all at once, the interface becomes unpredictable. One change can break something in a totally different place. It becomes scary to work with, let alone extend.”
So how do you fight back?
“The most important thing is prevention,” says Anton. This includes strong engineering culture, rigorous code reviews, architectural discussions before coding begins, and informal technical meetups.
At Hiveon, Anton’s team follows a two-track strategy: daily micro-maintenance and targeted cleanups. “We allocate 10–15% of every task to pay down minor debt — remove unused code, rewrite shortcuts, add tests. Over a few months, that pays off massively — without needing to stop development for dramatic ‘refactoring Tuesdays.’”
In parallel, they maintain a dedicated backlog of technical debt, with priorities based on how much each item affects system stability. “Every 2–3 sprints, we take several of those into the main development track — library migrations, refactoring complex logic, extracting shared utilities.”
To justify time spent on technical work, Anton emphasizes the importance of speaking the business’s language. “You have to show the numbers — how much time is spent fixing bugs, how long setup takes, how often releases are delayed. If you can say that an infrastructure change will speed up deployment 4x — that’s an argument the business understands. Switching from Webpack to Vite gave us exactly that.”
Sometimes, the cost of neglected debt becomes clear only after damage is done. Anton recalls one case where users experienced lags when scrolling long lists. “We thought it was a backend issue, but it turned out each list item was a heavy React component with multiple effects and subscriptions. One change would trigger mass re-renders.”
Using React Profiler, the team identified the bottleneck and introduced virtualization with react-window, also moving heavy logic out of the components. The results were immediate: smooth scrolling even on low-end devices, and CPU usage dropped 5–7x.
Today, Hiveon’s frontend team applies strict quality standards: consistent code style, naming conventions, and structural rules. Tools like ESLint, Prettier, and Stylelint are enforced via Git hooks. They also use AI assistants during code reviews, particularly for architectural consistency.
Code review, Anton stresses, isn’t just a formality: “Every merge request is reviewed by at least two developers. It boosts code quality and ensures knowledge sharing. Before large tasks, we also run short design sessions — they often prevent mistakes that could later become debt.”
For teams already buried in technical debt, Anton offers practical advice: “First — don’t panic. Don’t try to rewrite the whole thing. Start with a health check: get the team together, list the most painful issues, create a rough backlog, and prioritize.”
“Second — move in small steps. Fixing even one issue can motivate the team. The business will see that this invisible work brings real value. After that, it’s all about consistency: methodical cleanup, team support, and regular alignment with the product. That’s all it takes.”
Find a Home-Based Business to Start-Up >>> Hundreds of Business Listings.
