Share Article
News
An enterprise-grade implementation of ChatGPT, a natural language interface to our data warehouse and an automated generator of logic trees. These are just some of the ways Nous Group is using generative AI to support its consulting work with clients around the world.
Over the past year, expert consultants from across our international network have been exploring the capabilities of AI (artificial intelligence) to understand how it can improve productivity and drive better outcomes for our clients, their communities and our consultant teams.
Nous Principal Rob Ewin, lead of our AI working group, said: “AI brings exciting opportunities to reexamine our approach to a growing array of processes. It has the potential to free up our clients and our people so they can spend more time and energy on higher-order problem solving that delivers better outcomes and bigger impact.
“Across sectors, there’s great enthusiasm to unearth and harness AI’s value. This is exciting but it must be supported by systems that safeguard the accuracy of output and the security of information.”
Creating the right conditions for innovation and development
Our enterprise-grade implementation of ChatGPT, known as NousGPT, runs on customised Microsoft technologies and operates as a closed loop system, enabling our teams to safely experiment and develop their skills in AI.
Nous Principal Sophie O’Connor, our Project Excellence lead, said: “In building NousGPT, we’ve created a safe environment for our consultants and corporate teams to develop their confidence while getting a practical understanding of how AI can deliver greater value.
“Investing in our people and our capabilities is a key focus at Nous. AI is moving fast – how we adopt and adapt now will set us up for bigger success down the track.”
Going beyond the chat interface
Nous’ experimentation with generative AI has included prototyping two internal productivity tools, known as DawnGPT and Diana.
Nous Principal David Diviny, our Chief Data Scientist, said: “Generative AI can be so much more than chat bots. Thinking beyond a chat interface has the potential to scale the productivity impact of generative AI.
“This requires a deep understanding of a user’s tasks and how generative AI can overcome limiters to the ideal experience of that task.”
Our first productivity tool is DawnGPT, a natural language interface to the Data Assets Warehouse at Nous (DAWN), which brings together proprietary data and public information to facilitate rapid analysis. DAWN includes data on topics including population, human services, economic indicators, tertiary education and job advertisements.
DawnGPT provides an intuitive interface that seamlessly translates natural language into structured query language (SQL), providing consultants with a means to access data and an avenue to hone their SQL skills. The data results can then be visualised interactively or as a PowerPoint and a narrative can be generated to complement the results.
Our second productivity tool, Diana, generates logic trees for structured problem solving that breaks down a question or problem into its constituent parts. It is essentially a visual representation of structured thinking, with each branch representing a subset or component of the main question.
David said: “Diana uses generative AI to give our consulting teams a head start on projects. It is not the end solution, but the beginning of a creative problem-solving process.
“The possibilities for applications like these are endless and it’s an exciting journey to be on.”
All use of AI at Nous is guided by clear principles to ensure its use is safe, secure and generates genuine value that enables our teams to push the thinking even further and deliver higher quality work for our clients.