It’s rare to create something unique in the tech sector – something with no groundwork or infrastructure to guide you. 

But when they started developing the Offboarding Assistant, that’s precisely what Annabelle Sauve, Katie Van Luven and Dr. Raj Singh’s lab in the Dept. of Cognitive Science at Carleton University did.  

We launched the Offboarding Assistant in June 2023. Think of it as a digital expert that “knows” provincial employment standards legislation. If you need to terminate an employee, it’ll ask you a few questions and calculate your termination pay obligations. In a few clicks, you’re compliant. 
While the final product is intuitive and straightforward, the journey here wasn’t. The idea formed a few years ago when we began hearing that a disproportionate number of small businesses were sanctioned under their provincial labour laws for terminations. We examined it, and the numbers told a striking story: Ontario alone saw 11 718 ESA violation investigations in the 2022-23 fiscal year.    

In most cases, businesses aren’t at fault – they have to cut through a jungle of legalese to understand provincial employment legislation.  

So how did we make these complex labour laws digestible? This was Katie Van Luven’s first hurdle. She began by translating each province and territory’s employment legislation into a logical code using a programming language called Prolog.   

“Basically, I came up with all the logic, the computations, the calculations – all the work we’re trying to alleviate for an employer or HR professional,” she said.  

Here, Katie’s education – a mix of computer science and linguistics – was invaluable; she’s completing a Ph.D. in cognitive science at Carleton University.  

“Cognitive science is an interesting interdisciplinary field,” she explained. “Writing in Prolog is very similar to how you write natural language logical forms in first-order logic and semantics – which is a subfield of linguistics.”  

But a logical translation was only one piece of the puzzle. Katie and the team then needed a way to determine the relevant rules. They settled on a dialogue method – the tool would ask the user a set of questions and apply the logic to prove compliance.  

This iteration involved connecting the Prolog code to C# – and that’s when Annabelle Sauve joined the project. She was studying cognitive science at Carleton and working as a research assistant when she started helping with debugging. Now, she’s responsible for most of the Offboarding Assistant’s programming.  

The dialogue method still posed a challenge. To cover all the legislation, the team needed to pose an excess of questions; one early version asked the user over 30 questions.  

It proved too cumbersome, so the team struck a balance – they added some assumptions to each calculation.  They called it the “stay out of jail” (SOOJ) approach. An employer may pay more than required, but never less. They’ll always be compliant. 

For Annabelle, the user experience was critical – the tool had to suit busy business owners.  

“We originally had this super long, multi-page assistant,” she said. “And when we made the switch to the version where we only ask two or three questions, I thought, ‘OK, we’ve got this. It’s really going to work.’”  

Katie and Annabelle didn’t just face challenges with the product; they often struggled with the legislation itself.  

“I remember we had to call the province like the legislation province and be like, ‘Hey, can you tell me about this rule?’” Annabelle said. “Because not all the information is as available as you would think.”  

Katie admits that many of the issues she called about were specific – but that reinforced the project’s importance. If provincial legislation was unclear, how could businesses remain compliant?  

“Sometimes I would call multiple times, and each time I called, someone would tell me something a little bit different,” she said. “And this obviously isn’t true because the law didn’t just change in between calls.” 

There was no room for ambiguity in the Offboarding Assistant – employers and employees need the correct calculation.  

These hurdles stemmed from one issue: this is the first time anyone in Canada has undertaken a project like this.  

“From the beginning, it was so difficult to find the best way of doing things because it’s something that no one has ever done before,” Annabelle said. “There was no example of it – we just had to figure it out on our own.”  

But there was inspiration in the challenge, too.  

“A big part of what makes me want to do this is that it’s being done for the first time,” Katie said. “I love the challenge of solving a totally new problem – especially a problem that’s going to help employees get the pay they deserve.”  

“In the tech landscape we’re in, so many things are just repeated,” Annabelle added. “It feels good to be part of something novel and special.” 

Now that the Offboarding Assistant is available, we’re seeing the benefits of homegrown innovation – this was developed in Canada for Canadian businesses like yours. Discover exactly how the tool can help you stay compliant.  


Leave a Reply