Towards an Era of AI in Data Driven Decision Support - SensusQ's Winning Mind

Jaan Übi

Abstract

The two constraints facing the dual-usage decision support automation agent, when compared to yesterday, have remained unchanged. Firstly, in an adversarial situation, both parties have equal amounts of one resource, namely, time. Secondly, regarding cognitive overload, once data collection is in place, the fact remains that, as we take OSINT (Open Source Intelligence, e.g., X/Twitter or Telegram) clues, for example, in established worst-case situations, critical information has taken approximately two weeks to surface in the public domain. This processing stigmergy, evident in the inverse of information cascades from a distributed collective filtering process, owes a lot to the large number of participants scanning the information. But what if the discovery is even more complex, stemming from the combinations with other, importantly, proprietary "intelligence's" - signals intelligence, human intelligence, image intelligence, geo intelligence, etc.? How to create a data-driven decision support tool for such a purpose?

SensusQ puts forth an AI solution that, having its primary constituents named after the functions in the human brain, solves the problem by addressing the central, top-down, information processing requirements, and by discovering, bottom-up, the "unknown-unknowns" (or things we don't know that we don't know). Winning Mind has fast, slow, memory and executive AI streams that adhere to the constraints and provide the solution.

About Speaker

Jaan Übi earned his PhD in data mining from TalTech in 2014. He then became a Fulbright scholar and a postdoc at Salisbury University in Maryland, USA, from 2014 to 2016, and a research fellow at the University of Tartu from 2016 to 2017. He thereafter held senior data scientist and team lead roles in the banking industry, at Swedbank and Nordea, working in Tallinn from 2016 to 2020 and in Stockholm from 2020 to 2024. He specialized on customer facing and financial crime prevention models. He was also a conduit between the University of Tartu and Swedbank, kickstarting the AI cooperation between the institutions and helping to facilitate the creation of a productive industry-academia AI ecosystem. Today, Jaan is the Head of AI and Lead Data Scientist at SensusQ.