Supporting qualitative research projects through the safe adoption of AI
Verian
We accelerated the process of summarising interviews on some studies, freeing up more time for in-depth analysis.
Background
Verian carries out social research to inform policy design and decision-making for governments around the world. Analysts constantly conduct interviews, manually summarising transcripts to codify key points, which can be labour-intensive. Verian wanted to see how AI could speed this up to enable analysts to spend more time on research design and in-depth analysis.
“This cutting-edge new technology, powered by Generative AI, has the potential not only to drive efficiency and cost savings, but also to fundamentally improve how qualitative analysis is done in future.”
Tom Nixon, Managing Director
Faculty
Solution
Using large language models (LLMs) to summarise interview transcripts based on defined topics, analysts can spend more time interpreting results. The tool also automatically suggests further topics to give analysts more ideas to supplement the original research. Analysts are able to edit the summaries – so there’s always a human in the loop – with links back to the original transcript so all model outputs can be double-checked.
Impact
The tool has changed the way analysts work, freeing up more time to apply their expertise now that summaries can be created automatically in minutes rather than hours. As the overall process is quicker without reducing quality, this also enables Verian to bid more competitively for new work and conduct more detailed analysis.
“We’re really excited by the possibilities that are now available through the appropriate, secure and responsible application of artificial intelligence, alongside our existing leading research methodologies.”