The system of AI agents, called Multi-Agent AI Framework, includes a tool called Unstructured Data Agent, which helps insight teams understand qualitative feedback.

Unstructured Data Agent also combines AI features including thoughtfulness scoring, focused on evaluating the depth, relevance and quality of open-ended responses, and AI probing, which creates follow-up questions to qualitative responses.

Other features include AI summarisation for open-text data and insight reals, which turns open-ends and video recordings into video narratives.

Future agents – including on planning and development, analysis, and reporting and knowledge – will work to provide further AI-led support functions into the research process to help automate time-consuming tasks.

Researchers will also be able to adopt a “super agent”, which is an agentic model capable of managing the end-to-end research process, with human oversight.

Rival Group is also inviting brands to explore and help shape the next wave of its AI research innovation by participating in its AI early access programme.

Jennifer Reid, co-chief executive and chief methodologist at Rival Group, said that the tools were “not about automation for automation’s sake” and would provide users with “more time for strategic thinking, more confidence in their insights, and more compelling ways to deliver value to the business”.