Taigawise
In 2024, when EU legislation around sustainability reporting suddenly became more complex and demanding, Taigawise CEO, Hanna Liappis, already leading in the field of double materiality analyses (DMAs), set out to meet surging client demand. She called on the tech team at NordHero to build her a bot.
Using generative AI, large language models, and machine learning, the NordHero team began bringing the new Materiality Analysis Tool (MAT) to life, designing it to automate the most challenging and time consuming parts of performing corporate DMAs.
You might wonder why, if MAT is powered by publicly available large language models and generative AI, Taigawise didn’t use those directly? The simple answer is security.
By building their own personalised tool within the AWS framework, Taigawise has been able to keep all client data fully protected and secure. “This is especially important when dealing with sensitive sustainability and financial information,” says Liappis.
AWS also ensures that all data is handled compliantly. Teemu Niiranen, who led the Nordhero team on the technical side, explains, “By building MAT on AWS, we were able to ensure that Taigawise’s customer data would be handled securely and compliantly.” In fact, AWS has a compliance program that respects even the strictest governmental security requirements.
“Working with a partner like Taigawise, who already had a clear concept in mind, made it so much easier to get going,” recalls NordHero CEO Pekka Malmirae. “Because we knew the benefits and the use case from the start.”
However, midway through the project, a surprise shift in EU legislation left the team temporarily flat footed. “It was a bit of a rollercoaster,” Liappis recalls. “When the EU’s omnibus legislation evolved, we had to rethink our original use case as many of our potential clients no longer had to carry out compulsory DMAs.” Thanks to MAT’s in-built adaptability, the tool is now capable of evolving into other areas of sustainability reporting, like carbon foot-printing and biodiversity impact studies, which are also relevant for their clients as they navigate the shifting sands of EU legislation.
“The flexibility of AWS and the multi-tenant architecture is what allowed for these expanded capabilities, while still ensuring MAT would be able to serve different clients with specific needs, and even move through different language requirements,” says Malmirae.
Malmirae sees the ability of generative AI to integrate with company information systems as being a game-changer for Taigawise and many other clients, citing its ability to understand natural language, analyse documents, and summarise data effectively.
During the project, NordHero was in Las Vegas at the AWS re:Invent conference, where Pekka Malmirae and Teemu Niiranen stumbled into a workshop on multi-tenant Generative AI approaches – just the kind of cutting-edge expertise that could be directly applied to MAT’s development. Luckily their new knowledge didn’t stay in Vegas. They brought it back with them to Finland, ensuring that the project stayed at the forefront of AI advancements.
“It’s always great to work with experts who really know their stuff and are leaders in their field,” says Liappis. “The team of Finnish engineers at NordHero are to the point, yet so friendly and nice.”
But how friendly is MAT? The tool is still in its early phase, but there is one quirky aspect that Hanna and her team are still developing. His personality.
“From the beginning, we wanted MAT to behave like more than just a sustainability reporting tool,” Liappis affirms. “He should behave like an extension of our team. That means be responsive, helpful, and aligned with Taigawise’s core values of courage, joy, cooperation, and expertise.”
“It’s like teaching a toddler to talk,” Malmirae says. “You set the tone, and the child, in this case MAT, learns to keep that personality throughout its interactions.”
This focus on personality is not just whimsy. The goal is to craft a user experience that mirrors the way Taigawise works with its clients, which makes good business sense. So, in the future, when clients interact directly with MAT, he will do more than just churn out expert responses. Liappis hopes he will adopt a tone that’s even a little joyful.
Just like the Taigawise team.
Thinking about an in-house AI-driven project of your own? Click here to connect with NordHero team.
