Artificial intelligence goes from cleaning to happiness in companies

When opening the door to the gym’s bathroom, the person who is going to shower triggers an alert to the cleaning staff – it’s time to clean. The data generated, in turn, feeds an algorithm, which uses artificial intelligence to organize the workers’ routine and measure their satisfaction.

This day-to-day rationalization of the daily life of companies, which seemed a distant reality not long ago, is the bet of Brazilian startups, taking advantage of the spread of 5G and the cheaper sensors.

“Cleaning has been done the same way for decades”, says Leandro Simões, CEO of Evolv, which offers solutions for the building maintenance market.

According to him, cleaning is “a very important process, but one that is so neglected that nobody monitors it properly”. Therefore, the startup works with customers to define the right metrics to track.

“It is a market that in Brazil turns more or less R$ 100 billion per year”, he says. “There’s a lot of money on the table to earn and a lot of wasted water to save.”

The executive saw the opportunity after more than a decade in the telecommunications sector, when he worked in the acquisition of Telefônica by Vivo and in an antenna startup, where he had contact with investors such as GP Investments and the Blackstone fund.

At Tractian, the business opportunity was perceived by a young and practically inexperienced team.

“Most of the first employees met there at USP’s Escola de Engenharia de São Carlos”, says João Vitor Granzotti, responsible for the startup’s data area — he completed the course last year.

Tractian has created hardware and software capable of predicting, with high accuracy, when an industrial machine is about to break, anticipating maintenance processes and preventing the production line from being interrupted.

“Tractian’s idea is to empower maintenance teams in industrial processes”, says Granzotti.

By monitoring and recording in real time information such as vibration frequency, temperature and power grid parameters, the startup’s algorithm detects sensitive changes, identifies patterns and alerts customers to potential problems in a user-friendly interface.

The sensors are capable of monitoring up to 60 types of machines. Customers range from automobile industries to a shrimp farm, says Granzotti.

Founded during the pandemic, the startup already serves customers in countries such as Argentina, Chile, the United States and Mexico.

In Fiter’s case, the optimization target is not a process or a machine, but the employees themselves, through a “happiness index”.

The methodology was developed by CEO Sergio Amad, an executive and professor with more than ten years of experience in human resources and neuropsychology.

The calculation is based on the answers to eight questions, which are partly repeated and partly renewed, once a month. Workers respond if they want.

The questions are multiple choice and start with statements such as “I am satisfied with my performance” and “I see that my profile is compatible with the role”, with which the employee can agree, disagree or be neutral.

With the responses of most employees of a company, the algorithm is able to identify patterns and notice when someone has changed their mood in relation to work.

According to Amad, the main results are a reduction in turnover and an objective measure to know if any employee is close to having burnout, a situation that has become more frequent during the pandemic.

“This happiness measurement gives us the opportunity to pay attention to those people we were a little distracted to”, explains Toni Gandra, founder of the EcoFit gym, which is a Fiter customer.

He says he has already reversed two potential layoffs thanks to the analysis the service offers. Weverson Alves, training supervisor at Live One Trade, says that the system serves not only to understand employee happiness, but “also what the company can offer them.”

Businesses with artificial intelligence attract investors

The efficiency of SaaS companies (software as a service) that use artificial intelligence to improve processes of other companies, has been drawing the attention of investors.

Last month, Goldman Sachs led a BRL 625 million investment in unico, a Brazilian digital ID unicorn that is now valued at $2.6 billion. With more cash on hand, the company has invested in research and development of proprietary technologies.

One of the fronts, in partnership with the Federal University of Paraná, invests in periocular biometrics (analysis of data from the eye region), with the aim of improving recognition and mitigating algorithmic biases.

“Many of the facial recognition technologies on the market today were developed in northern countries, based on Caucasian and Asian faces,” says the company.

“Our team invests in the continuous improvement of this proprietary technology, precisely with a focus on improving the experience and access for all people.”