MatchingID recruiting with artificial intelligence
| Reference number | |
| Coordinator | MatchingID AB - MatchingID |
| Funding from Vinnova | SEK 300 000 |
| Project duration | April 2017 - November 2017 |
| Status | Completed |
| Venture | Innovative Startups |
| Call | Innovativa startups fas 1 Våren 2017 |
Important results from the project
Our digital service will match employers and talent interest, as accurate as possible while eliminating name, picture, gender and age. The purpose of this project was to find out how AI/machine learning can make this process progressive and efficient. Now we have a concrete plan for how to make machine better than man, at achieving on point talent sourcing, with an holistic approach but without human prejudice. We have formatted data equivalent of the data a headhunter needs to match in a way so that it can be used by a self learning, self improving intelligent algorithm an AI-headhunter.
Expected long term effects
With very promising results we have mapped our how and started producing our own AI/machine learning that can analyse and learn from data and behaviours to make relevant conclusions about what makes a good match, based on both the talents and employers needs and wishes. We have met tremendous interest from universities, HR-departments, recruitment agencies, talent networks and sought after talent. Our tech team confirms, together with our other stakeholders, that we can achieve a larger visible supply of talent and improved diversity in Swedish companies.
Approach and implementation
Now we have determined what the least viable data need is to make the match a human headhunter does but better. We have a technical plan, a cost overview and a team to take us all the way with the AI-algorithm. We have through in depth interviews and wireframe tests confirmed that the business idea and business model as well as willingness-to-pay is viable in the market. To achieve a high quality algorithm we now need larger amounts of data and more time, so that the neural network can learn and finetune match quality. We have a database in place.