Our annual Roadshows are a great opportunity for us to share the latest SystemsLink news in person. It turns out that this year’s theme, “Energy Management and Artificial Intelligence” truly resonated with the attendees.
In case you couldn’t attend, we were able to premiere something we have been working very hard on: The next generation of our flagship energy management software “Daisy” (the clue is in the name!).
Daisy’s debut allowed us to introduce our audiences to the concepts of generative, agentic and predictive AI. Crucially, we could bring forth the idea that these solutions can only work well if they are powered by a domain-trained Large Language Model (LLM); This allows humans and machines to communicate easily.
But what do we mean by trained? Well-established LLMs ensure Generative AI such as Chat GPT and Gemini are wholly intuitive in their use of natural languages like English. However, to meet the very particular demands of energy and sustainability managers, those LLMs need to know a lot more about our industry.
That’s why we are fine-tuning our proprietary “energy management” LLM with enormous volumes of data on weather and climate change, carbon factors, published tariffs, historical consumption patterns and the like. We are training our AI tools to make sense of all this and help us spot issues, track performance and predict the future.
In time, our clients will be able to harness these capabilities – combined with their own segregated supply contract, estate, consumption and capital improvement project data – to predict, respond and accomplish the day-to-day cost management and long-term carbon reduction demands more efficiently.
Like all our software, the all new, cloud-native “Daisy” is highly resilient and the data within it highly secure. However, in the era where AI has become mainstream, additional guardrails are essential.
For “Daisy” this means we progressively develop a “human-in-the-loop” autonomous decision-making capability, for example, how and when to contact a supplier about an invalid invoice. In collaboration with our clients, we will determine together the parameters governing what the AI may and may not do.
This is a very exciting time for energy and sustainability managers – if only we could manage our climate as intelligently as “Daisy” manages data!
