It is no secret that artificial intelligence (AI) has evolved exponentially in recent years. From smart cars and automated robots to simpler solutions such as a personal home assistant that can turn up the A/C, AI has become an integral part of our everyday life.
Many believe the goal should be to make AI increasingly smarter and powerful, but what if the purpose of the technology shifted towards becoming more sustainable? According to Nature Communications, when AI is applied to contexts like the energy sector, it can play a vital role in achieving the Sustainable Development Goals (SDGs) identified by the United Nations, such as increasing affordable and clean energy (SDG 7).
When entrepreneur Uljan Sharka realized there was this need for sustainable technologies, and especially AI, he founded iGenius, an AI company that develops augmented intelligence solutions for business data. The company is well known for its seamless AI-advisor, crystal, which we have worked with for a couple of years. “Innovation and sustainability cannot but go hand-in-hand and this is why we work hard as a native AI company to innovate and disrupt the B2B data industry to make it efficient and accessible”, mentions the CEO. This is an objective that Enel is wholeheartedly aligned with.
Meet crystal, your new go-to advisor
crystal is a conversational AI that auto-classifies business requests, based on underlying information, to understand users’ needs. Sharka explains: “The technology uses a business knowledge graph, also known as the AI brain, to connect the dots automatically, giving power to the users as they define the use case and what KPIs they want to measure”. He adds that “once this connection is completed, crystal uses metadata to identify data sources and act as a unified vehicle data link, which means that it will retrieve the information in real time, and then virtually normalise it and stream it back to the user interface in a format that is readable and comprehensive”.
This AI is currently in its second software version, which has allowed it to be lighter, faster and smarter than before as well as to easily connect to any open channel. What once took three to six months to integrate into a company’s data set, now only takes minutes with the updated version. According to Sharka, “The beauty of this system is that it can connect to a company’s data set as a virtual overarching layer that sits on top of your infrastructure without impacting it, making it more efficient for any business”. Besides this, crystal democratises the learning experience through its user-centric interface, which makes it easy for the less digitally-literate to adopt and understand it.
As Sharka sums up, “crystal is much more than just another conversational AI: we are now using the technology for smarter data integration and improved transfer learning capabilities”. With the global energy transition in mind, the CEO believes that “although innovation and technology can be powerful enablers, they also come with their own price tag in terms of resources and overall societal, environmental and efficiency impact”. This is why crystal’s second version is now aimed at being more sustainable in terms of technology, adoption, and overall energy efficiency.
A crystal-clear sustainable future
One of the biggest factors that defines crystal as a novel sustainable AI is its ability to “cut computing power by half”, claims Sharka. In traditional AI, users need to provide hundreds or sometimes thousands of question examples to train the API; crystal auto-classifies data in order to skip this step. Sharka reveals that “instead of having to re-train the AI over and over again on each topic that crystal is serving to users, we have developed a technology that is able to recycle that knowledge and generate fewer data sets, thus reducing the overall computing power”. In addition, crystal leverages data through its streaming functions instead of replicating it, which makes it lighter in terms of performance, storage and overall power consumption.