ETIA: Causal AI for Data-Driven Insights and Optimal Decision Making

Objectif

Medical doctors strive to heal people. Subscription-based companies aim to retain their customers and increase profits. Car manufacturers try to increase the safety of their cars. To achieve their goals, scientists and businesses make decisions based on their expected outcomes; these are determined by the causal relations in the respective domain. We will develop ETIA, the first Automated Causal Discovery software platform embedded within a Causal AI Agent. ETIA, given the clients’ data, automatically discovers, explains, and visualizes the plausible causal relationships, optimizes decisions, and performs root-cause analysis. As an AI Agent, ETIA will be able to converse with the user in natural language. ETIA provides value to customers by leading them to optimal decisions and data-driven insights. Major research parts for ETIA were performed during an ERC consolidator grant and industry-funded grants, while a prototype was built as part of an ERC PoC. The project will take the platform to TRL 6, prepare it for the market and for attracting VC and angel investment. The total addressable market is part of prescriptive analytics, AutoML, and Agent markets. Competitors include Data Robot, H2O, and Dataiku but their products focus on predictive analytics which is a surrogate and substitute for true causal analytics. The ETIA team has 20 years of experience in terms of research in causal discovery. The PI of the project is a seasoned entrepreneur having founded the award-winning JADBio company as a University of Crete spin off and developed the JADBio SaaS product with more than 10000 registered users and 20 paying customers. For the market validation of ETIA, 4 tech companies have already offered their support, namely Honda Research Europe, Huawei Technologies (Ireland), Phenomena+, and JADBio. ETIA intends to scale up globally to achieve a significant financial and scientific impact.