How a Waterloo Arts student is building better tools for research

A psychology study that once took five or six weeks to build can now be created in less than an hour. Jacob Oliveira, a PhD student in the School of Psychology, developed the new tool by aiming to solve his own research challenge. Now his solution has grown into a venture shaped in part by Waterloo’s culture of experimentation and entrepreneurship.

“I came here to study people, not to build software,” he says. “But that’s what the research kept demanding of me, until I decided to fix that.”

His answer was JOLIVE, a no-code platform for building and running behavioural studies. Researchers can use it to create surveys, video and image-based studies, classic psychological tasks, multi-participant chat rooms, and experiments involving AI agents. For researchers studying AI interaction, JOLIVE also offers optional integrated AI chat blocks that can be configured for a study. Studies are built through a visual interface, while JOLIVE handles complex processes such as counterbalancing and condition ordering. Researchers can preview their study designs, make changes immediately and collaborate with multiple editors on a single study.

That efficiency reshaped the quality and quantity of research Oliveira could conduct. In a single semester, he used JOLIVE to run eight studies involving 840 participants. He is now involved in roughly 15 studies at once, with the platform allowing him to manage that scale of work.

“Without JOLIVE, I’d be left debugging 15 projects. It’s impossible, and not a good use of any researcher’s time,” he says.

JOLIVE is designed for behavioural research broadly. Its optional AI study tools have also placed Oliveira in the middle of one of psychology’s most urgent conversations: how people understand and relate to artificial intelligence. His own research explores AI trust, perceptions of ownership, and environmental impact, questions he believes are too often left to technical fields alone.

“There’s this technological gap,” he says. “A lot of the AI research is being done by people in highly technical fields, which is fantastic, but psychology and other social science fields ask different questions. We ask more human questions.”

Word travelled fast. As other Waterloo researchers started asking to use the platform, JOLIVE outgrew its original purpose as one student’s solution.

Oliveira's advisor, Dr. Daniel Smilek, says the platform is already proving its value to behavioural researchers.

"JOLIVE is an outstanding tool for creating state-of-the-art psychological studies examining the intersection of human psychology and artificial intelligence. We have used the platform successfully to examine the factors that drive human trust in AI systems and their outputs. The platform is stable, reliable, and highly useful."

JOLIVE has become a fully incorporated business, with Oliveira building a team that includes development and marketing expertise. A Waterloo Computer Science co-op student is also helping shape the next version of the software.

Oliveira points to Waterloo’s innovative culture as one reason the idea has grown. It is a place, he says, that is collaborative and experimental by nature, where students are encouraged to test ideas across disciplinary boundaries. Faculty met his unconventional psychology PhD project with curiosity and support.

“When I started building this, I didn’t run into a lot of people saying, ‘You have your work to do. This isn’t worth your time,’” he says. “My advisor and other faculty were willing to say, ‘That’s interesting.’ They were also honest about the product, offering constructive advice when things were not quite right and giving praise when it hit the mark.”

Oliveira is now working to bring JOLIVE to researchers well beyond Waterloo and to develop it as a teaching tool through which students can learn experimental design by building studies themselves.

His goal, in the end, is a simple one: “I want people to be able to ask the questions they want to ask without technology standing in their way. Let’s get back to the research instead of the coding.”