• Written by Muhammed Abdulla NC, CEO

A new human resources system is launched, promising to streamline employee data and improve decision making, but it quickly becomes clear that the system is not being used as intended, and the anticipated benefits are not materialising, a shift supervisor checks the maintenance log and finds that the system is not even being updated regularly, leading to inaccurate information and poor decisions.

Automating a process nobody mapped

One of the most common mistakes when applying AI to human resources is automating a process without first mapping out the existing workflow, this can lead to a system that is not tailored to the specific needs of the organisation, and can even automate inefficiencies and biases that exist in the current process, a category manager compares two suppliers and finds that the system is not taking into account key factors that are crucial to the decision-making process.

The consequences of this mistake can be far-reaching, leading to a system that is not trusted by employees, and does not provide the anticipated benefits, it is wrong to assume that AI can simply be applied to a process without first understanding the intricacies of that process, and the specific needs of the organisation, most people get wrong that automation is a simple matter of replacing human labour with machines, but it is much more complex than that.

Treating policy library as complete when it is not

Another mistake is treating the policy library as complete when it is not, this can lead to a system that is not able to provide accurate information, and can even provide misleading information, the policy library is the foundation of the system, and it is essential that it is complete and up-to-date, a human resource manager reviews the policy library and finds that it is missing key policies and procedures.

This mistake can have serious consequences, leading to non-compliance with regulations, and even legal issues, it is essential that the policy library is regularly reviewed and updated, to ensure that it is complete and accurate, this requires a significant amount of time and resources, but it is essential to the success of the system.

Giving the capability only to the analyst who already had it

A common mistake is giving the capability to only the analyst who already had it, this can lead to a system that is not accessible to all employees, and can even create a bottleneck, where only one person has the ability to access and analyse the data, a data analyst is the only one who can access the system, and this creates a bottleneck, where all requests have to go through them.

This mistake can limit the benefits of the system, and can even create resentment among employees who do not have access to the system, it is essential that the system is accessible to all employees who need it, and that training is provided to ensure that they are able to use it effectively, this requires a significant amount of planning and resources, but it is essential to the success of the system.

Declaring success against a metric with no baseline

Declaring success against a metric with no baseline is another common mistake, this can lead to a false sense of accomplishment, and can even mask underlying issues, a manager declares that the system has been a success, based on a metric that has no baseline, and this creates a false sense of security.

This mistake can have serious consequences, leading to a lack of accountability, and even a lack of improvement, it is essential that metrics are established before the system is launched, and that regular reviews are conducted to ensure that the system is meeting its intended goals, this requires a significant amount of planning and resources, but it is essential to the success of the system, for more information on how to establish metrics and measure success, please contact us.

When AI is the wrong choice

Finally, it is essential to note that AI is not always the right choice, there are certain situations where AI is not appropriate, and can even be counterproductive, for example, in situations where the data is incomplete or inaccurate, or where the process is highly complex and nuanced, in these situations, it is better to rely on human judgement and expertise, rather than relying on a system that may not be able to fully understand the complexities of the situation.

This requires a significant amount of careful consideration, and a thorough understanding of the limitations of AI, it is wrong to assume that AI is always the answer, and that it can be applied to any situation, this is not the case, and it is essential that the limitations of AI are carefully considered, before making a decision, for more information on how to determine whether AI is the right choice, please consult with an expert.