There are hundreds of factors that improve our quality of life. In fact, many of them have been analyzed scientifically. It has now been proven that change in our activity levels, diet, alcohol and nicotine consumption, housing situation, and work-related stress levels make us more or less happy in our lives.
WorkWell is an application based around those principles. It makes use of a body of academic research that was translated into Artificial Intelligence and Machine Learning algorithms, which help measure almost two hundred different factors of human lives and compare them to the average levels. The algorithm then offers recommendations on how we can make positive changes in our lives.
Praxis Workwell Limited is a British research and development company that created the concept of the application. The project was to answer two questions: what makes us feel fulfilled, and happy (the so-called “quality of life), and how it affects our productivity at work (Human Capital Productivity).
We have implemented this concept and made a solution available in the cloud. The project was carried out entirely by our .NET team – creating WorkWell was a big challenge for FINGO developers. They have worked with the awareness that this app can really make the lives of many people better.
In the second stage, our developers implemented the project for a British government organization. Then, new Covid-19 questions had to be matched to the algorithms. It was essential to understand how the global situation influenced the employees’ well-being and how they could be supported while maintaining their productivity.
The WorkWell app is mainly aimed at employers who can anonymously ask their employees questions about various aspects of life. The tool also suggests what the employers can change to improve the overall happiness of their staff.
Creating appropriate questions for the questionnaire, required 5 000 paid survey participants to build a knowledge base for algorithms. It was necessary to determine statistical answers.
This very thorough analysis leads to positive changes at work, which can directly impact every company’s efficiency. Having satisfied and relaxed employees leads to less absenteeism, better productivity, and increased revenues.
To build a reliable and efficient ML model, we used scikit learn library. It is a high-level API written in Python. The original idea was to use deep neural networks, but we decided to use the scikit learn library due to specific requirements. It fulfilled its task perfectly – also under heavy load.
We managed to develop a questionnaire with 200 questions that can be answered within 15 minutes. The questionnaire contained conditional questions, which required a good approach to user experience and data presentation.
The back-end was developed in .NET while the front-end technology of choice was React.