New Research Project: Bringing AI risk intelligence to corporate treasuries, supported by STEP funding
What would a sudden rise in energy prices or a sharp currency swing do to next quarter’s margins?
For many mid-sized companies, answering that question can take weeks and often requires external advisers. Effective exposure measurement tools have mainly been available to large financial institutions with dedicated risk teams and significant budgets. As a result, energy-intensive businesses and those with international supply chains often respond to shocks only after they occur. Our new research and development project addresses this challenge and has been selected for a grant under the Emilia-Romagna Region’s STEP 2025 call, co-funded by the European Union.
About the Project
The project will deliver a new Axyon IRIS module that provides corporate treasuries with advance visibility. Beginning with a company’s data and documents, it will generate a verifiable report identifying business exposures. Then, users will be guided through what-if analyses on currency, energy, commodity, and interest rate shocks, while IRIS’s predictive signals will estimate the likelihood of each event in the coming months. Treasury teams will know their exposure and the probability of shocks before market changes, allowing time for informed hedging and financial decisions. Institutional clients will also benefit from enhanced market regime detection, stress-tested portfolio construction, and a systematic process for evaluating new AI model families.
The innovation lies in integrating these elements into a single decision chain. Predictive signals, market context, and an AI interface that reads company documents typically operate separately. In this project, they work together, enabling users to move from their own documents to a clear view of exposure and likely scenarios within one workflow. The AI reasoning layer is guided by structured data, quantitative models, and a complete audit trail, ensuring every figure in a report is traceable to its source. This traceability allows treasury teams to confidently present results to their boards and ensures alignment with EU principles for trustworthy AI.
Partners
“Learning” is one of our core values, and academic research remains central to our innovation. We will collaborate with the AIRI research centre at the University of Modena and Reggio Emilia (UniMORE) and conduct large-scale training and validation on CINECA’s supercomputing infrastructure to ensure know-how, IP, and talent remain in the region. The Region’s €580,475 grant supports a total investment of €1.1 million in the project. We thank the Region for its support and look forward to sharing our progress in the coming months.
To learn more about the technology behind our solutions, check our dedicated section: