From predicting consumer behaviours to pre-empting driving conditions, artificial intelligence will help unlock new insights for both businesses and individuals. As a recruitment services provider, CPL is pioneering the use of machine learning to help match the right candidates to the right job opportunities.
We previously heard from Aidan Keenan: A Trailblazing Data Ninja Building AI Apps, but today I caught up with Kevin Sweeney, IT Director at Dublin-based CPL, about how it plans to make the most of AI.
In the last 12 months, we’ve seen an explosion of new AI products aimed at the recruitment industry, primarily focused on matching candidates and predicting success rates. There have been lots of conversations around the idea that AI will replace consultants but I really don’t see that happening. Finding the right candidate is only the first part of the recruitment journey – building relationships with clients and candidates is also key and that will always need a human touch.
We want to make short-listing candidates easier for our 350 consultants. We receive around 40,000 CVs every month, and that’s an awful lot of information for them to go through. To help our consultants be more productive, we’ve built an AI-enabled app, based on Einstein and Heroku, which uses machine learning to predict the probability of a candidate’s success.
Within 12 hours of receiving a CV, we can process it via the app and provide consultants with a list of the most highly rated candidates. It helps us react faster and boosts our competitive advantage – recruitment is all about speed.
For machine learning to work, you need a wealth of data to analyse. Our app draws on eight years’ worth of candidate and CV data, such as education and career history, which has been captured in a standardised format in Salesforce. We’ve developed and trained a model, using machine learning and our historical data, where we can now input in a candidate’s information and output their probability of success.
You’ve got to have good base data to get accurate results. It’s got to be clean, consistent, and comprehensive. It also helps to have a strong mathematical skillset within your organisation that understands the data, so you can verify outputs and confirm that your model is working correctly.
I think we’ll see massive developments in the use of AI in the next year or two, particularly when it comes to transactional placements, such as recruiting 50 people to steward a concert, that aren’t so much about the candidate’s skillset. There are loads of exciting opportunities in other sectors too – I’m a bit of a sports fanatic, so I’m really interested in how AI might be used in the future to predict a team’s performance.
And in the meantime, read this e-book on AI for CRM to find out how you can get started.