Most common mistakes when recruiting software developers/engineers
According to the current statistics, a poor recruitment process makes up to 50% of job candidates decline a job offer, while nearly 31% of hires decide to leave their newly acquired positions within the first six months. Whether those employees were indeed not a great match for the companies they have left or not, a high turnover ratio always means losses for an employer.
Recruitment, hiring and training procedures take time and money, while a lack of competent specialists generates expenses related to the delays of ongoing and future projects. The problem is especially acute in the IT sector with its high requirements for technical skills. Moreover, the more new people come into a company, the higher the risks of leakage of unique business knowledge. Finally, fast turnover weakens the morale of other staff members and builds a negative reputation on the market.
We asked a few of our Devopsbay team members what they think about issues with the right recruitment process.
Senior DevOps - Kuba and Łukasz
From your experience, what mistakes can you see when we are talking about Devops position?
Conducting a type of question / answer recruitment may not fully determine whether a given person fits the team. Oftentimes, someone may know the answers to all questions, but not quite coping with problem-solving and teamwork. I am a supporter of a rather open discussion, in which you can get more from the candidate in terms of technology and whether he is a skilled person. Skillset can be enlarged, but characters won't chang
From your perspective, what are the DevOps competencies needed in a startup and what in corporations?
I don't know if it matters - I have never encountered any differences during such recruitment. You check the technical stack and / or soft skills in terms of what you have in the project rather.
Can you tell us also - What counts when you are recruiting a DevOps Engineer?
I always check whether the candidate is interested in the topic and whether he is expanding his knowledge, in most cases, candidates do not understand the mechanisms of action, and only recreate what they saw
Other Javascript/ Generally developers - Ola
What mistakes can be made when recruiting a developer?
It seems to me that the biggest and fundamental mistake is to approach recruitment in IT in the same way as to recruitment in other industries. The IT market is very specific. In this market, the candidates are dictating the terms. The first mistake that we can make right from the start is the lack of specific information in the first message to the candidate.
Good specialists received dozens of messages with invitations to contact during the week and the recruiter must make sure that his message stands out and encourages conversation
The next mistake is the lack of good knowledge of the position we are recruiting for.
Sure - you can always say that we are not sure, and we don't want to mislead the candidate and you will return to him with confirmed feedback, but this should be very sporadic and only used in exceptional circumstances. The well-known and well-worn phrases in advertisements such as "due to the dynamic development of the company" will also certainly not help in recruiting candidates.
When finally we manage to “catch” the contact with the candidate, firstly, we should simply approach the candidate in a human way. We shouldn't treat the candidate as an opportunity to achieve the "KPIs" but as a person whom we want to help find the right place that will utilize his competence and experience.
Of course, it is impossible to treat everyone uniquely and individually,but a huge mistake is the use of scripts when a candidate needs to be approached individually and conducting recruitment in a schematic manner.
MLOps - Damian
MLOps projects required a lot of knowledge and an understanding of the process. Looking from the recruiting site what mistakes do you see?
There is a problem with MLOps. Some people understand MLOps as deploying a product or doing something in Kubernetes and many more. They are all right in a certain sense.
The biggest problem lies with the decision-maker. They don't quite feel what they actually want to do with MLOps.
They don't know who they want to employ. So they employ someone who has the best skills, but they do not necessarily translate into the task that he is to perform in a given project.
Companies have a problem with determining what to do, how to describe the processes, how to do analyzes and then do not know how to implement it all correctly. Identifying the needs of MLops - is the greatest challenge. If this is done wrong, the entire recruitment process will be incorrect
What do you think, are there some most important MLOps competencies needed in startup projects and in corporations?
There is no difference. In both, you need the transparency of the process and its repeatability,
If you have some models and effects, there comes a time when you verify the profitability and return from implementing MLOps. Business is all about hard data that helps you make a good decision. MLOps works the same for startups and corpo
The basis is the MLops added values. that is, how the model behaves and how long it will be up-to-date. When MLOps meets business, it has to provide it with some predictivity based on data. Again, this works the same for startups and corporations
