UNIT.City — місце, де люди працюють... КРАЩЕ! Обирай свій простір просто зараз 👉
Наталя ХандусенкоWork
27 June 2025, 09:00
2025-06-27
Is the skill of writing prompts for AI already a guarantee of an offer or just a buzzword? We asked developers and team leaders how necessary it is today
“This is already a must-have, there are already cases where engineers are asked to write online coding prompts during interviews,” is one of the comments we saw on LinkedIn. It was written by Software Developer Viktor Svertoka after discussing such cases with colleagues during a meeting at the DOU Day IT conference. So we decided to ask team leaders whether they practice such a task, and vice versa — we asked developers whether they have encountered such a requirement. In addition, we monitored vacancies on jobs.dou to check how often employers require candidates to have such a skill. Here is what our small study showed.
“This is already a must-have, there are already cases where engineers are asked to write online coding prompts during interviews,” is one of the comments we saw on LinkedIn. It was written by Software Developer Viktor Svertoka after discussing such cases with colleagues during a meeting at the DOU Day IT conference. So we decided to ask team leaders whether they practice such a task, and vice versa — we asked developers whether they have encountered such a requirement. In addition, we monitored vacancies on jobs.dou to check how often employers require candidates to have such a skill. Here is what our small study showed.
What the developers say
dev.ua received responses from 23 developers, most of whom said they had not encountered the requirement to write AI prompts during technical interviews. Only one IT specialist wrote about such an experience. At the same time, the responses showed that some companies are against this approach, while others, on the contrary, ask to use AI tools during test tasks. Such an ambiguous attitude of companies to AI leads to doubts, especially among juniors, whether it is worth informing IT managers and recruiters about your skills and experience in involving AI tools in the process of writing code.
Among all the IT professionals interviewed, only Front-End Developer Dmytro Tymoshenko said that he had encountered such a requirement: "Indeed, I had such an experience at an interview at a startup specializing in training language models and creating web interfaces for their use. Since this is a startup, development speed was very important, so they actively used modern tools with integrated AI models to improve and speed up their work. One of the tasks at the interview was to write a prompt for generating tests, as well as a prompt for creating mock data. This helped to check whether I could quickly and clearly formulate tasks for a neural network to get the desired results."
"Speeding up your work" is one of the main factors why companies need IT professionals with skills in using AI tools.
There is a boom in such specialists in the US right now — every fourth IT vacancy will include a requirement to work with AI tools, writes Business Insider. We recently wrote that in Silicon Valley, companies are increasingly posting vacancies that require the use of AI tools specifically to accelerate product release , including Visa, Reddit, Udacity, the streaming service Tubi, and others.
One of the respondents said that there have been cases where companies have asked to use AI to perform tests.
"While I was looking for a job, I had not yet encountered live coding, technical interviews were conducted orally, at most they asked me to explain how the code, which had been prepared in advance, would behave. Sometimes before the technical interview"They gave me a test task and asked me to use AI to the maximum in my work. They said that they actively use artificial intelligence in their work and they would like to see how the candidate works with such tools," Software Engineer Mykyta Krukovsky told dev.ua.
At the same time, IT professionals say that companies are often against AI tools for programming, especially in interviews .
"This is not desirable, especially when the company does not want to publish the code of its project, its work. Although I have seen a couple of companies in vacancies that, on the contrary, encourage the use of AI," noted PHP Developer Volodymyr Marakhovsky .
On jobs.dou you can indeed find vacancies that require AI tools, but there are not many of them yet. Here are a few examples:
You can see vacancies that mention “interest in AI tools” or “passion for AI” as an additional plus. For example, Python Developer in the Zakupivli.Pro team (the largest tender platform in Ukraine in the Prozorro system) and Intern Python Developer in the Dutch outsourcing company HYS Enterprise.
They are also looking for specialists who already have experience working with AI tools, but this is not the main requirement yet, but only an additional one. For example, Python Tech Lead at Ukrainian IT company PLANEKS and Junior/Middle Python Developer at Astronics CSC.
In addition, there are companies that require experience working with specific AI tools . For example, the requirement "ChatGPT / Copilot / Codeium for rapid code writing and refactoring" for the Senior Python Engineer (AI) vacancy at the service IT company Kozak Group.
However, despite the emergence of such vacancies on the Ukrainian IT market, IT professionals sometimes do not understand how open they can be with HR and recruiters about their ability to work with AI coding tools, due to the ambiguous attitude of employers towards the use of this technology. Especially when it comes to juniors who are at the start of their IT careers.
For example, this was mentioned by a novice developer, whose name we are not mentioning because he is looking for a job.
"Actually, I'm not a fan of vibe coding. I'm more interested in learning technologies and coding manually, but having AI act as my mentor. But due to certain circumstances, the idea has developed that if a person uses AI and HR finds out about it during an interview, then a lot of condemnation often falls on the person, it's an immediate minus, etc. Because of this, it's not entirely clear how open and honest I should be in terms of telling whether I use AI."
What do you think, Timlids?
The questions to the team leaders consisted of several parts:
if you practice, what exactly does the task contain;
if not, they were asked to explain why;
and whether they plan to add such a task in the future.
To summarize the team leaders' responses, none of them practice writing code prompts as an interview task. And they explained why they are against this approach. At the same time, some are considering adding such a task in the future.
1. Lack of time. The technical part of the interview usually takes 1 to 1.5 hours, and the higher the level of the candidate, the more technical topics to discuss and the less time left for similar secondary questions.
2. Policy of IT companies and customers. Most companies and customers still prohibit the use of AI for code analysis and direct development (live debugging, AI IDE, live autocomplete/copilot, etc.), due to fear of code compromise or potential theft.
3. Hard skills > prompt engineering. I am convinced that technical knowledge and the ability to independently write good code are fundamental skills that will help in understanding how to build not only a quality system, but also a quality prompt if necessary.
From my observations, if a person has a fairly strong knowledge of their main stack, then writing a good prompt is an easier task for them than for a person with strong prompt writing skills, but without an understanding of the technical component. For the latter, it will take at least not one prompt, but several, in order to first understand what exactly they need and how to correctly achieve the goal of the query.
But if there is a conditional customer who will give carte blanche to use AI tools and prioritize their use on his project (I have also encountered this twice among colleagues), then I would not mind adding questions of this type because in this case they will start to have more meaning and greater weight in everyday work.
Personally, I don't ask candidates to write prompts during interviews. The main reason is that I'm mostly looking for DevOps specialists, and DevOps is not just a set of tools or scripts, it's a philosophy. It's about understanding what the customer wants... even when the customer doesn't understand it yet.
I always give a “plus” to those candidates who honestly say: “I haven’t worked with this, but I know how to find a solution — through ChatGPT, Google, or Stack Overflow.” This shows the main thing — the person is not afraid to learn.
My interviews focus not only on the technical stack - it can be memorized. It is important for me to understand whether a person can think, adapt, and is able to learn new things. If they can learn, then they are most likely already using AI - not as a calculator, but as a teacher. Not just: "Here is a question - give the answer," but with a desire to understand why exactly.
In summary: the issue of using AI in interviews is not simple. But for me, what matters is not whether the candidate can write prompts, but how they work with information, whether they can think critically, and whether they learn with AI, rather than just copying the result.
I don't see a need for this for the roles I'm hiring for yet, but the market is changing rapidly, so I don't rule out this possibility in the future if the role of AI tools in DevOps changes dramatically.
We had different requirements, including completing a test task in a few days at a convenient time, or live coding in Laravel or pure PHP via screen sharing, but not writing prompts using AI.
Not yet. We're thinking about where to implement AI. The ability to write prompts in live coding is unlikely, because I evaluate what a person can do as an engineer.
I haven't hired anyone to my team in the last 3 years, the team is stable, without any hiring or firings.
If I were doing an interview today, I wouldn't ask to write a prompt. This is not a hard skill that is of primary need and is relatively easy to acquire; there are much more important questions that I would be interested in asking.
No, I don't practice this and I don't plan to. In my opinion, it makes no sense, it's very easy to understand how a person will do prompts (if necessary) by listening to how the candidate communicates and answers technical questions.
That is, it is worth paying attention to how a person explains technical aspects; if this skill is at a good level, then he will be able to write requests in the conditional ChatGPT.
But, in my opinion, asking a question about the ability to write prompts makes absolutely no sense, and it would be unwise to judge technical candidates based on this.
I'm not asking about prompts yet, but in the near future, it will be an additional question.
In general, I am currently developing a full circle process for working with projects using AI and LLM. I will rather teach than search and ask, hard skills are much more important for now.
GitHub CEO Thomas Domke recently said that in the future, testing skills in working with AI tools will become mandatory in interviews.
He says the company is already thinking about how to use AI in interviews. Domke notes that if you want to get a job at a tech company in the near future, you'll likely be asked to demonstrate your skills in prompting and using AI agents.
“Because the goal of the future engineer is no longer to start everything from scratch. The goal is to combine their hinting skills and agents, open source libraries to solve this problem much faster than they could two or three years ago,” says the CEO of GitHub.
As a summary
The involvement of AI tools as assistants in programming in Ukrainian companies is only at the beginning: employers are cautiously indicating this requirement, disguising it as "it would be a plus"; team leaders do not yet see the point in testing these skills, and accordingly, IT professionals rarely encounter a task at interviews where it is necessary to use artificial intelligence.
“This leads to severe degradation and laziness.” Ukrainian programmers shared the secrets of effective vibe coding and the “side effects” of using AI when creating code
Vibe coding is becoming increasingly popular in IT — programmers simply ask AI to perform certain tasks and do not delve into the actual code that the chatbot offers.
Як нейромережі бачать вільну та незалежну Україну? Тест dev.ua
Нейронні мережі для генерації зображень бачать світ по-своєму, їхню логіку зрозуміти часом зовсім неможливо. Але таки хочеться. На честь Дня Незалежності України редакція dev.ua вирішила провести невеликий експеримент.
Ми задали чотирьом різним нейронним мережам п’ять однакових запитів: «прапор України», «День Незалежності України», «український Крим», «перемога України» та «українці». Отриманими результатами ми ділимося з вами нижче.
У TikTok тепер можна генерувати фон за допомогою нейромережі. Ми протестували її та ділимося результатами
У TikTok з’явилася нова функція «Розумний фон». З її допомогою як фон для тіктоків можна підставляти згенеровані нейромережею зображення. Редакція dev.ua протестувала цю технологію і ділиться своїми враженнями.
