By Anastasiia Lushyna
Welcome to the new episode of our interview series! As you know, artificial intelligence has revolutionized multiple industries, and its usage keeps increasing. The software development field is no exception; however, the application of AI has caused plenty of debates. While AI offers multiple benefits, there are also certain threats.
To clarify the situation, we decided to interview Maksym Voloshyn, our Development Team Lead. Maksym has been working at Attekmi for 3 years, and his skills and knowledge are invaluable to our progress. Keep reading to discover what an experienced Team Lead thinks about the usage of AI in software development.
It is a pleasure to welcome you to our interview series – thanks for joining us! Before we start, can you please tell us a bit more about your career in AdTech?
Well, AdTech became a true discovery for me. Obviously, I knew that advertising was a powerful tool for any business; however, I had no idea about its technical aspects. I didn’t know how it worked. Everything changed three years ago when I joined the Attekmi team as a Backend Developer. I entered the world of RTB requests, SSPs, DSPs, millisecond speed, and enormous amounts of data. That sparked my thirst for knowledge. I wanted to learn not only technical nuances but also the entire business context.
Step by step, I delved deeper into the technical details. I started to suggest improvements and take greater responsibility. And one year ago, I became a Tech Team Lead. Now my key task is not only to develop the product but also to support the team's growth, help make technical decisions, and build the internal culture.
Yes, AdTech is a complex sphere, but only if there are no experienced and supportive teammates next to you. At Attekmi, that’s not a problem – we all help each other. At the same time, AdTech is a very interesting industry, and I keep evolving in this direction every day.
Nowadays, it seems that AI is being integrated into workflows across nearly every industry, streamlining processes – even in software development, where it is widely used for code writing. What’s your perspective on the growing role of AI, and do you currently use it in your work?
Actually, to answer this question, let’s focus on the role of AI in code writing. Yes, AI can speed up the process and reduce the number of errors – we are all humans, and even the most experienced developers are still prone to mistakes. AI can also analyze and optimize code and make coding itself more accessible for those who are only entering the development sphere.
However, you should always remember that AI can make mistakes as well, so the generated code may be of poor quality. Besides, there is a problem of the ownership of the code written by AI – there is still a lack of proper regulations and standards. Finally, the biggest risk is that if you use proprietary code or any other sensitive information in your prompt, it can be used to train the model further. As a result, your secrets can be disclosed.
At the same time, AI can still be an ally when used carefully, and my team and I use it as well. I guess the adoption of AI in software development will only increase since the automation aspect is a great benefit. However, if you use it for code writing, you should check every generated piece of code thoroughly and navigate the legal landscape. I would also suggest not relying on AI too often. If you use it for code writing continuously, your own skills will never improve.
Are there any development-related tasks you would never ask AI to do? Why? And which tasks are actually safe to complete with artificial intelligence?
Let’s start with the good news – the tasks that you can actually use AI for without risk. So, first of all, AI can assist you with project roadmapping and planning. For instance, you can ask it to estimate the project scope, suggest the most relevant technology for this specific project, and identify potential risks. Secondly, AI can be an effective supportive tool during the design and architecture stages – e.g., for API design or scalability considerations. Besides, AI can generate for you test cases, which will speed up the process of software testing.
As for the tasks I would never ask AI to perform… I would not recommend using AI for debugging across multiple architectural layers – it can solve isolated errors, but when it comes to real-world environments, it is rather useless. Besides, AI cannot conduct in-depth security audits – the technology is powerful, but it lacks human expertise.
What is the most significant advantage of AI in software development for you? Can you think of anything applicable specifically to the AdTech industry?
Basically, the most significant benefit of AI is that it can automate routine tasks and accelerate the development process. However, as for me, it is the most useful in terms of project management. AI can help with time estimation, resource allocation, task scheduling, etc. While certain aspects are automated, you can focus on more important things that definitely require human supervision.
As for the AdTech industry, the greatest thing you can do with artificial intelligence is to equip your solution with AI capabilities to ensure precise targeting.
Have you ever integrated AI into software development processes? If yes, was there any specific case when the technology performed even better than you expected?
Yes, we actively integrate AI into our development processes. It all started with basic things: generation of simple scripts that are used by our admins and DevOps experts, writing complex SQL requests and coding hints via Copilot or ChatGPT… However, after some time, we started considering AI as a full-scale team consultant that can deal with a wider range of tasks.
One of the specific cases that I can recall (and it turned out to be a pleasant surprise for me) happened when we were testing the auto-generation of SQL requests to the analytical base. We spent pretty much time on creating complex requests to ClickHouse and MySQL for the dashboards in Grafana. We tried doing this on our own, but the requests were not optimized. In turn, AI not only provided us with a specific result, but also optimized the part of the logic that we used to write manually. This saved our time and provided us with suggestions on how to review the entire approach to data aggregation.
Another interesting thing is using AI for generating mocks and fake data in testing. It used to be a pretty boring task, but now we have automated it almost entirely.
In general, I consider AI not a replacement for a developer but a tool that allows for freeing up some time for more creative and strategic tasks. The main thing is to ask the right questions – you need to learn how to write prompts in the right way. If you are interested in this, I would suggest taking the course on how to communicate with AI.
One of the most common concerns is whether AI could replace software engineers. What do you think about it? Is this a real threat to developers?
Well, AI is not able to fully replace software developers – its performance still needs to be monitored, the generated code or whatever else must be reviewed, and so on. It is a tool, not a virtual team member that can work fully independently. However, engineers should learn how to use AI in software development to gain a competitive advantage. More and more companies are starting to leverage AI, and this is not likely to change in the near future. Therefore, AI will not replace developers, but it may affect the way they work.
At the same time, AI can start posing a threat to entry-level professionals at some point. For instance, a business may entrust AI with code generation under the supervision of a middle or senior expert – this way, there may be no need for a junior developer in the team. What to do about it? Improve your skills continuously, gain new knowledge, and, obviously, learn how to use AI in your job.
Do you think keeping an eye on AI evolution is essential for software developers? If yes, why, and what is the best way to do this?
Yes, understanding what AI can do at the moment and what opportunities it may offer in the future is crucial for engineers – this ensures greater flexibility, just like in the case of any other trend. I would recommend monitoring industry forums and news sources – that is the easiest way to stay on top. When possible, attend relevant conferences. AI is an extremely popular topic nowadays, so the majority of events include it in their agenda.
As a team lead, do you somehow ensure that your team stays on top of AI innovations?
Well, if I find some interesting and relevant news, I always share it with my team, and they do the same for me. We all want to be aware of the latest technology advancements – that is one of our secret efficiency techniques at Attekmi.
In your opinion, how will AI impact software development in the future? Will there be any specific applications in the AdTech industry? And what about potential threats? Should we expect new concerns to emerge?
I expect more enhanced capabilities in terms of code writing. That is the most time-consuming part of the development process, so most likely, the majority of advancements will impact specifically code generation. Obviously, AI will also support a wider range of languages and frameworks.
In terms of software development in the AdTech sphere, it will adopt the same advancements as other industries. However, since AI plays a significant role in programmatic advertising, more sophisticated analytics and targeting capabilities may enter the stage, and AdTech platforms will have to integrate them to remain competitive.
As for concerns, the most significant ones are already here, and they can evolve. For instance, regulations outlining the ownership of AI-generated code may finally appear, which can complicate the work of those developers who tend to rely on AI.
As the Development Team Lead at Attekmi, could you share your top 3 or more AI lifehacks for developers in 2025?
The most important thing that you should remember is that overreliance on AI is not the way to go. Something may go wrong, and you will not notice this. You may face unpleasant legal consequences. Your skills will not progress. Actually, there are a lot of potential issues. So, if you decide to use AI in software development, do this responsibly – review everything it generates and never use sensitive information. If you are a lead, be an example for your team. Educate them on how to leverage AI and promote a culture of responsible AI usage. As for other tips, here they are:
Experiment with different tools. In this way, you will be able to select a solution that will meet your needs.
Align AI tools with your standards. If you decide to leverage AI for writing code, make sure that it adheres to your company’s coding standards. This will ensure the code’s consistency.
Keep your prompts specific. This helps AI tools deliver more relevant answers. For instance, you may specify such details as libraries, frameworks, programming languages, and so on. Be detailed – that is the key to success.
Document the usage of AI. In this way, the development process will become more transparent, while you will also collect insights on how to use AI more effectively in the future.
Learn continuously. Otherwise, you may miss some important advancements in AI technology and become unable to use it to its full potential.
New episodes of our interview series are on the way, so keep an eye on our blog updates to miss nothing! And if you are ready to start using one of our services or ad exchange solutions, just contact us!
