Looking for Industrial Partners: AI Agents for Knowledge-Intensive Process Automation

The project envisions a next-generation agentic AI framework that enables organizations to automate complex, knowledge-intensive processes starting from unstructured and normative sources such as documents, regulations, and informal practices, directly targeting real-world Apply AI scenarios. By combining large language models with multi-agent systems, symbolic AI, advanced planning, and verification techniques, the system delivers increased autonomy, robustness, and reliability of AI agents, in line with the call’s expected outcomes. One key focus is on ensuring controllability, trustworthiness, and compliance of AI agents with organizational, legal, and ethical constraints. A key ambition is to lower the barrier to AI adoption for SMEs and public organizations, which strongly aligns with the Apply AI objective of accelerating uptake beyond large technology players.

Nice-to-have profiles for industrial partners:
- organizations with knowledge-intensive processes that remain largely manual or semi-automated;
- sectors where work is governed by regulations, standards, or internal policies rather than fixed data pipelines;
- companies or public bodies interested in automating compliance-heavy, exception-rich, or evolving workflows;
- willingness to provide representative use cases, domain expertise, and qualitative feedback rather than large datasets;
- interest in trustworthy, explainable, and controllable AI rather than purely performance-driven automation;
- SMEs, mid-caps, public organizations, or service providers operating in regulated or complex organizational environments.

We are particularly interested in partners that can help us identify and refine concrete usage scenarios where AI agents can support or partially automate knowledge-intensive organizational processes, serving as demonstrators for the project’s methodological and technological results.