SSE transforms audits with tailor-made GenAI tools

For businesses in highly regulated industries, missing even one legal requirement or failing to identify a potential risk can be costly. That’s why they rely on teams of experts to carefully audit their operating policies and processes. But the real value these teams bring comes from their expertise and insights - not from the time they spend processing huge volumes of documents and data.

So when the Risk, Innovation and Analytics team at energy provider SSE saw a demonstration of PwC’s generative AI (GenAI) capabilities, they immediately recognised the potential benefits. Instead of painstakingly reviewing hundreds of pages of documents for 60-plus audits a year, they realised they could use GenAI to identify key information, allowing them to focus on more strategic, higher-level cognitive work. Working closely with us, SSE tested our GenAI tool - and the results have been promising.

Tailoring the tool

“Rather than spending days, weeks, going through policy documents, we get our summaries quicker now,” says Diana Cherepanova, SSE's Risk, Innovation and Analytics Lead. “We can focus on the actual problems within the audit, and we can focus more on the root cause analysis of the issues.”

For Cherepanova’s team, the most time consuming part of preparing for an audit is reviewing unstructured data held in text file documents rather than tables, such as supplier contracts, to see how well they align with structured data, such as purchasing activities. Working with PwC to test and customise the GenAI tool, SSE wanted to be able to identify the exact document and page number it refers to with every response it gives.

“This helped to validate the results and increase confidence in the outputs,” Cherepanova says. “Auditors talk to other auditors and say, ‘You can trust this tool. It’s great.’” By customising the tool to meet their specific needs, SSE is leading the way in the adoption of new technologies, while carefully managing risks.

Realising the value

SSE can rely on the GenAI outputs because the tool uses internal knowledge bases created specifically for different types of audits. This means auditors can spend less time searching for information and focus more on delivering insights. “Asking one question takes a few seconds, and going through documents takes a lot of time,” says Cherepanova. “And that time you can now spend on something more interesting.”

“SSE was doing great stuff around data analytics, but it was all around structured data,” says Karl Saunt, a partner with PwC’s Technology Data and Analytics team. “There was all this unstructured data they weren’t able to shine a light on because they couldn’t access it and they couldn’t consume it.”

Thanks to its transformative AI experiment, SSE now has that ability. “GenAI helps us to get information from unstructured data quicker and in a way that we were never able to get it before,” says Cherepanova. “Everyone is excited to test it, so we are planning on expanding the user base. Our next step is to teach people how to create prompts in the best way so they will get the correct answers. I hope that this will help us to discover new ways we can use the tool.”