After they raised an investment of $2.4M, just two weeks later they received a new investment, signaling that the startup is on the right track. The $ 12M round was led by Point72 Ventures, with participation from Sands Capital Ventures, Amadeus Capital Partners, Passion Capital, and Entrepreneur First.
PolyAI is “one very deep tech company” that deals with the design of a dialogue system. Specifically, the PolyAI platform is created so that AI agents can conduct complex conversations with customers within customer support. Nikola shared the motives for swimming in these stemmed waters from the competition point of view:
Too many chatbot companies have died trying to find favorable software licensing agreements years before their technology was ready for wide-scale deployment. Our main motive is to prove that these technologies can have a very bright future. After a short stroll, we decided to specialize in one domain of our dialogue systems platform—specifically, customer service. We want full alignment between PolyAI and contact centers.
So far, vertical integration has proven to be a very successful strategy for AI companies and it is the best way to fast-track the development of our machine learning platform. Otherwise, if you are working on a horizontal solution, you are clashing with giants like Google and Amazon, which makes it difficult to build an independent company in the long term.
In that sense, PolyAI systems are different from voice assistants which are developed by large technology companies. Their agents are specialized in specific domains, such as reservations or technical support:
Assistants like Siri and Alexa follow a different methodology—they try to answer a wide range of questions and it is imperative that they never make serious mistakes because their reputation is at stake. On the other hand, our assistants do not try to simulate general intelligence, but rather aim to provide reliable answers within a very narrowly specialized domain. They do not lose the context because they are based on a mathematically consistent formalism that gives them memory and the ability to model the context of the conversation.
On a question how they will use the funds they received, Nikola says:
We plan to use some of our funds to buy or invest in contact centers, to prove over the next two years that our platform can improve the productivity of our agents. We believe that we will be able to secure sufficient resources to develop technology and grow our team (which currently includes ten leading AI researchers in the field).
From Belgrade’s Mathematical Grammar School to the University of Cambridge and Apple Siri
Nikola finished Mathematical Grammar School in Belgrade, a prestigious national high school, after which he received a scholarship at the University of Cambridge. After his studies, Nikola got a job as the first engineer at VocallQ startup working to develop advanced conversational systems. After this startup became part of Apple through acquisition, Nikola spent two years in the Apple Siri team working on perfecting speech-understanding systems that are learned from human conversations. In parallel, he completed his PhD studies.
In 2017 all of this led him to bring together fellow scientists and engineers from the Cambridge Dialogue Systems Group, part of the Machine Intelligence Lab at the University of Cambridge. He slipped into entrepreneurship as, in his words, he wanted to put himself in a position where he controls the narrative and feels full responsibility for the product and company’s survival.
