Predicting careers
A company wanted insight in the decision-making process of people looking for a job. The practical question became: “What can we do to help our customers make the right choice?”
For us, this was primarily a matching issue. We needed to find out which qualities, characteristics and career plans of people connected to a combination of company profiles and job requirements. Custom algorithms enabled us to retrieve the required information.
What we did
Recommendation engine
Interaction and interface design
Web and mobile development
Finding the right job for the right person
Companies and organizations with job vacancies naturally look for the right person to do the job. Persons looking for a job tend to take a different view. They want a job that is right for them. So, we needed to discover which factors influence their decision to apply.
The project started with developing a so called minimum viable product (MVP), a basic version of the desired end product, meant to test a number of assumptions. This way of testing saves time and costs, since the final product is only developed when the MVP yields adequate results.
Note
In consultation with the client, we developed the recommendation engine with a number of possible uses in mind. The underlying matching system is also suitable for connecting products, services, people and many other items in a great variety of ways.
From custom algorithms to suitable advice
This specific type of matching started with finding out which characteristics in general are related to career choices. Think of hard and soft skills, education, experience, languages, ambitions, etc. By analyzing a large number of profiles and their interaction with other entities, including company profiles, retrieving the required information was on its way. This was done by developing custom algorithms that enable us to gather and process large amounts of possibly relevant data. We also compared career decisions by persons with similar profiles.
The results of the work of our data scientists enabled us (read: the client) to automatically generate individual career advice (specific job offers) with the highest probability that it matches the preferences of a specific job seeker.
Dynamic job recommendations
In a sense, this system updates itself. Once a certain job has been completed, it’s added to the profile of the individual, including a company rating and employer assessment. Other factors that influence one’s decision to apply for a certain job are also continually updated. Think of additional diplomas, newly acquired skills, age, experience or even a change in one’s career plans.
Validating the results
Part of the (updated) information is made available to the user. This helps work seekers gain more insight in their personal development and makes it easier for individuals to update their cv accordingly.
The matching process we developed delivers results that needed to be validated. In other words, we needed to make sure the job recommendations were relevant and attractive. Basically, this was done by checking actual applications, and acceptance by the employer. In addition, we asked the users what they thought of their personalised job offers. Simply because, despite all available cutting-edge technology: if you want to know something, sometimes it’s best to just ask.
The trick is to discover which data influence someone’s choice