Employees at webTactics have one less tedious task on their plates thanks to a recent collaboration between the recruitment marketing agency and Seneca Polytechnic researchers.

The time-consuming work of creating unique job postings — a small subset of the services webTactics provides — is now the domain of a custom generative AI app developed at Seneca.

And it’s already made the team at webTactics more efficient, according to the Toronto-based company’s vice-president of business development, Jeff Hurrell.

“I see it as a productivity tool. We're not looking to replace swaths of people with it or anything like that, but we are looking to enable them to be better at their jobs.”

He said part of what makes the task of creating job postings so time consuming is that each posting needs to be customized to each client’s preferred template, language and other specifications and then repeatedly refined.

Using data supplied by webTactics, an Applied Research team led by Dr. Amit Maraj, Professor, School of Software Design & Data Science, taught a custom version of OpenAI’s GPT-3.5 LLM to generate webTactics’ job postings.

Through a simple user interface, the user fills out fields with the job title, description, client-specific guidelines and any other information provided by the client, and the GPT creates the posting. If the user isn’t satisfied with the response, they can adjust the prompts.

The model doesn’t just produce job postings, the researchers concluded — it produces reliable, high-quality postings with minimal errors.

“Because of that, webTactics is not only saving money, but they're more efficient,” Mr. Maraj said. “They're getting more things done. They're taking on more contracts and being able to move forward as a business using tech enablement.”