When I started building my team, I was tasked with designing a technical exercise for hiring processes. It was our first mobile team, and we chose KMP technology. Few candidates had experience with KMP, and we didn't know if they would have the environment ready for a live coding challenge. In addition to the technical test, my engineering manager insisted on evaluating soft skills such as problem-solving, adaptability, creativity, collaboration, and communication.
I realized that evaluating soft skills with questions alone was not efficient. Anyone can use artificial intelligence tools to generate convincing responses with the words interviewers want to hear. Who hasn't made the mistake of hiring a developer with excellent technical skills but lacking in interpersonal skills? It has happened to me.
With this in mind, I designed an exercise that simulates the developer's daily work: a pair programming session with the candidate as the driver and the interviewer playing the role of a junior developer being guided. The candidate chooses a task from the backlog and guides the interviewer through its execution. The interviewer follows instructions, sometimes admits they don't understand or can't hear the candidate well, and sometimes disagrees with the proposed solutions.
The main goal is to observe how the candidate adapts to a new project with new technology, how they apply creativity to solve problems, what questions they ask, how they give instructions, how they handle frustration in the face of repeated failures, how they debug, and other relevant aspects.
From the start, we made it clear that there are no right or wrong results. We want to evaluate how the person interacts and integrates with the team. I have never seen two candidates approach a story in exactly the same way.
We have used this exercise with more than ten people, and several important observations emerged.
1. Candidates rarely asked what had already been implemented or how the existing functionality worked.
2. Even worse, they never asked to run the application to verify that the code works or to interact with the solution.
3. Most tried to demonstrate expertise by taking on the most complex tasks, which often led to frustration and writing unnecessary code.
4. Some froze, rejected suggestions to move forward, and did not engage in dialogue with the interviewer; it seemed like they were alone in the meeting.
5. When faced with communication or collaboration issues, they did not offer to take control of the meeting or write in the chat. Only one person lost control and interrupted, saying it was a waste of time because they were not willing to teach the basics.
The lessons I learned, which can be applied to any interview, are practical and simple. Never be afraid to ask obvious questions, such as asking to be shown how the solution works. Review the project structure and ask what is necessary. Find out if there are unit tests. Ask about the project's libraries. Check if there are business priorities or if you can implement what you feel comfortable with. After these steps, choose a task and, after reviewing the code, explain your plan to the team; this clarifies your thought process and demonstrates potential even if you don't finish on time. Communicate clearly and kindly, and if there are difficulties, consider alternative communication methods such as writing in the chat.
None of the candidates completed a task entirely. You might then ask if we managed to hire anyone. Yes, we have brought in excellent developers because completing a task was never the main goal for us. What we value most is bringing in people who live by the motto Alone we go fast, together we go further and who contribute real collaboration to the team.
At Q2BSTUDIO, we are a custom software and application development company specialized in tailored solutions and custom software that drive digital transformation. We offer artificial intelligence services and AI for businesses, AI agents, and Power BI solutions integrated into business intelligence services. We are also experts in cybersecurity and cloud architectures with AWS and Azure cloud services. Our approach combines technical expertise with good collaboration practices to deliver secure, scalable custom applications aligned with business priorities.
If you are looking for partners to develop custom software, implement artificial intelligence for businesses, deploy AI agents, integrate Power BI for advanced analytics, strengthen your cybersecurity, or migrate to AWS and Azure cloud services, at Q2BSTUDIO we offer consulting, development, and implementation with a focus on results and teamwork.
Evaluating candidates beyond code has helped us build strong teams capable of tackling complex and custom projects, and that is the philosophy we apply in every project with our clients.
Keywords custom applications custom software artificial intelligence cybersecurity AWS and Azure cloud services business intelligence services AI for businesses AI agents Power BI
