The goal for Abtran was set to be more selective and efficient during the hiring process in order to reduce call-centre employee rotation, and work towards decreasing their staff churn rate. Sophisticated machine-learning algorithms combined with Microsoft Power BI will enable them to achieve this.
Industry: Customer Service
Abtran expressed the desire to be able to change and optimise their recruitment practices. The goal was set to be more selective and efficient during the hiring process in order to reduce call-centre employee rotation, and work towards decreasing their staff churn rate. The high training cost associated with onboarding new staff was cited as a key factor Abtran was seeking to address and mitigate against as much as possible. It was estimated that new employees took up to two months to become fully productive so obviously it was in Abtran’s best interests to retain existing staff so that the two-month up-skilling timeframe for each individual employee didn’t go to waste (amounting to thousands and thousands of hours and a massive labour cost).
Abtran employ 2,300 staff across Ireland, the UK and Asia, and serve highly regulated organisations across Government, Transport, Utilities and Financial Service. Abtran work with these organisations to deliver excellence in customer service experience to their customers.
Codec designed an Azure solution to import existing on-premise employee data into an Azure SQL database. Machine learning predictive models were designed, tested, trained and ultimately deployed which would run complex algorithms on the data to determine likely outcomes. These outcomes were then relayed back to the Azure SQL database and ready to access from within the Power BI dashboard for Abtran management to review (and alter business practices in accordance with predictions where necessary).
Abtran are now in a position to access reports within Power BI that determine with an accuracy of 85% (and rising) the rate at which current employees are likely to leave. But machine learning goes far beyond that one metric and can also feed data on what gender, age range or department the exiting employee will be associated with. The ramifications of possessing this knowledge will enable Abtran to not only plan ahead for possible staff shortages, but also optimize their hiring process. The machine learning outcomes, for example, might inform Abtran that men between the ages of 21 and 25 from a particular department are churning at a higher rather than elsewhere. Thus Abtran could review that specific department’s hiring processes or simply hire people who are outside of this demographic and less likely to churn.
And the longer Abtran continues to train and use prediction algorithms, and the more data these algorithms can access, the more powerful and detailed the predictions will become, enabling Abtran to make more informed and proactive business decisions.
Machine learning is an emerging trend in the modern business landscape but is set to transform how businesses strategise and plan for the future. Almost like having Mystic Meg as your CEO!