We’re living in the epoch of AI. Day by day, you get more and more used to chatting with bots and assistants. Somewhere along the way, it’s easy to forget that behind every screen, there are still real humans building the tools you use. So today, for the very first time, meet flespi engineers – the people behind the code, the fixes, and the ideas.
We asked Evgenij Spitsyn and Nadzeya Mikhailava, our frontend part, a couple of questions about their roles and experience, what they enjoy most about their work, how they recover from a bug hunt, and their thoughts about AI, of course.
Evgenij Spitsyn
Frontend Engineer
"I joined the team almost ten years ago. Before that, I had been working on Wialon apps. In flespi, I’m a frontend developer responsible for the user interfaces and most of the frontend applications. This includes projects like the panel, Toolbox, and Setbox.
I started out as a PHP developer. At some point, I thought I’d move toward Python through frontend work – but somewhere along the way, I realized I actually liked frontend more, so I decided to stick with it. I still think Python is a great language, and maybe someday I’ll look more seriously at something like Go as well. For me, a language is just a tool to get things done, and I like what JavaScript lets me build.
What I love is building interfaces. I think developers can often be divided into two groups. Some love algorithms and pure logic, while others enjoy visual results. I’m definitely closer to the second group.
It's nasty that you can create interesting visual effects directly in code – animations, transitions, dynamic elements – often without relying on heavy graphics assets.
But, probably the same thing most frontend developers struggle with – browser inconsistencies and platform-specific bugs. Frontend evolves very quickly, but browsers sometimes lag behind, with their own quirks and limitations.
Frameworks help smooth over many of these issues, but there are still edge cases where you have to deal with strange browser behavior. I constantly run into issues, particularly with certain browser-platform combinations.
Here, I think AI tools are incredibly powerful. They speed up development, help with routine tasks, and allow developers to focus more on architecture and higher-level decisions. At the same time, they change the nature of programming. The amount of “manual” coding you do decreases, and the role of the developer becomes a bit closer to an architect than a builder.
However, AI still needs supervision. It can generate very nice-looking code that still contains logical problems. So, validation is essential. You still need to read the code, test it, and sometimes guide the AI while it’s working out the solution.
AI usually works within a fairly narrow context. Yes, it can spot many specific things in the code and keep track of them while solving a task. But it still lacks the broader understanding of the whole project that a human developer has.
Right now, AI is already very strong in frontend development because the ecosystem is well-documented and widely discussed in the community. AI can also suggest solutions you might not have considered. But if it disappeared tomorrow, I would probably feel both disappointed and relieved at the same time.
On one hand, it would mean we created something really powerful but couldn’t make it sustainable or affordable enough to keep it going.
On the other hand, maybe movies like Terminator or WALL-E were trying to warn us about something. Either machines end up running everything, or humans become so comfortable that we stop doing much at all.
And then one day we realize the planet is full of trash and we have to leave it behind and live somewhere else. That probably wouldn’t be the smartest way to use the resources we have here.
What else can I say about myself?.. I’m definitely a bit of a tech geek. I enjoy experimenting with smart home technologies and making my house a little more intelligent over time. I even run a local LLM at home for experiments. I also enjoy working with graphics and 3D modeling. I really like drones – both for photography and for FPV flying. Flying FPV drones gives you a sense of freedom.
...And freedom is a spirit of flespi. :)"
Nadzeya Mikhailava
Software Engineer
"Well, my story began in the QA department. And I thought I would surely continue working as a tester at flespi, but it quickly became clear that testers weren’t really part of the workflow. So I gradually started moving into development instead.
Initially, I worked more on the hardware side. I was involved in developing the compiler and working with device protocols. One of the first things I worked on was the Queclink protocol; we kept improving the PVM language and added new features to the compiler as long as protocols were implemented.
Over time, I ended up touching many different parts of the system – device-related work, backend components, and eventually the frontend as well. I’m also responsible for interacting with customers through technical support. So I probably became one of the few people in the team who has worked across several layers of the platform.
Right now, I’m mostly working on frontend tasks, but that doesn’t mean I only do frontend. If needed, I can switch back to other parts of the system. For example, recently, when the frontend work slowed down for a bit, I picked up some hardware-related tasks again.
PVM and protocols development is probably still the area where I feel the most confident. But with modern AI tools, the situation is changing quite a lot.
Today, you don’t necessarily need to know every language in extreme depth from the start. If you understand the fundamentals, you can quickly get up to speed with a new technology using AI assistance.
Of course, deep knowledge still gives you a big advantage. But learning something new is much faster now than it used to be. The whole learning paradigm is changing, and it’s still not entirely clear where it will lead.
I enjoy improving things and making them cleaner and more consistent, reducing the entropy in the system. Sometimes, I notice that several parameters or pieces of logic were implemented slightly differently just because they were added at different times. When I take the time to unify them and make everything more elegant, that feels really satisfying.
AI tools are fascinating but also a bit paradoxical. Sometimes they allow you to achieve 80% of the result with only 20% of the effort. You can generate a large amount of code very quickly – that would otherwise take a long time to write manually. But reaching the final 100% still takes work. You have to review the details, adjust things, fix edge cases, and understand what’s actually happening.
But it's about my work... Behind the scenes, a lot of my free time is naturally taken up by family life. Recently, I also felt a growing desire to do something creative with my hands – things like knitting or crafting. It’s something I enjoyed as a child, but never had enough time for later because of other responsibilities. And recently, I also started learning Argentine tango, which has been a fun new experience. :)
Yep, running helps a lot during a break at the office. When I get stuck on a problem and can’t find a solution, going for a run often helps reset my thinking. During the run, I stop actively forcing the solution, but somehow the answer appears on its own.
Regarding flespi, I should say there are many products in the world, but truly outstanding ones are rare.
In any market, you might find hundreds of similar options, but only a few of them are really well designed and thoughtfully built. I believe flespi is one of those unique things in its field. The team focuses on building the platform the right way rather than chasing trends or competitors. And I’m proud to be part of that."
So, do you feel you want to talk to the rest of the team?... ;) This is only a small part. We’ll introduce more of them in the upcoming posts. Stay tuned!