KALI
KALI – AI-driven module for optimized traffic categorization and rule-based filtering part of the KALEIDOSCOPE platform
Use Case:
Enhancing Distributed Denial of Service (DDoS) mitigation by integrating AI techniques to optimize rule generation reduce filtering overhead and support network scalability.
Outcome:
Three core technical challenges identified and addressed through feasibility planning; AI-supported traffic categorization and dynamic rule set optimization; Business model for AI module deployment in security platforms developed and tested; Achieved TRL-7 for the KALEIDOSCOPE architecture; Strategic evaluation of hardware limitations and filtering architectures.
Ecosystem Support:
Supported by the StairwAI program through business mentoring and technical guidance, facilitating agile development of a feasibility plan, risk mitigation strategy, and tailored AI adoption roadmap.
AI Relevance:
The project demonstrates the integration of AI in network security for SMEs through: I) automated filtering rule optimization using AI; II) clear alignment with platform-level architectures (KALEIDOSCOPE); III) realistic feasibility evaluation and business impact modeling; IV) focus on scalable and adaptable AI modules for cybersecurity applications.
Summary:
Level7 S.r.l. undertook the KALI project to enhance its KALEIDOSCOPE platform, a DDoS mitigation solution that defends IT infrastructures against large-scale cyberattacks. The central goal was to develop an AI-based module capable of improving traffic filtering performance while operating within the hardware limitations typical of SME environments. The team focused on three interrelated technical challenges: accurate traffic categorization, generation of a comprehensive rule set for blocking malicious flows, and creating a reduced rule set that could fit within the filtering capacities of network devices. After evaluating various strategies—including the use of more powerful but costly hardware—it was concluded that simply upgrading infrastructure was not a sustainable or effective approach. Instead, the emphasis was placed on smart rule optimization and AI-driven architectures. A thorough feasibility plan and business model were developed during the StairwAI Support Program. This included defining the commercial promise of KALI, identifying target users, evaluating key risks, and outlining go-to-market strategies. Agile methodologies were employed to refine the AI adoption strategy through iterative feedback with mentors and stakeholders. Despite some minor delays, the project reached its planned deliverables and demonstrated the potential for future growth through collaborative R&I initiatives. The business model canvas revealed high market interest and underscored the urgency of DDoS solutions in an increasingly digital and attack-prone world. Post-project plans include integrating the KALI module into future Horizon Europe consortia and expanding its capabilities through custom or layered filtering architectures. The KALI project successfully validated a novel AI use case for enhancing cybersecurity resilience in SMEs, setting a solid foundation for further industrialization and deployment.

