Jan Clement

I check software for vulnerabilities – before attackers find them.

My focus is on the security of AI integrations and LLM-based systems. Companies are now integrating Large Language Models into their products, often without knowing what new attack vectors this creates. Prompt injection, tool abuse in agentic AI systems, uncontrolled data exfiltration via AI interfaces – these are not theoretical scenarios, but real vulnerabilities that I regularly find and fix in client projects.

What distinguishes me as a consultant: I am not a pure AI consultant nor a pure pentester. My background includes 25+ years of development experience – from electronics and firmware, through software architecture, to the application layer. Anyone who really wants to test AI-powered systems for security needs to understand how software works at every level. I bring this understanding.

My services include:

– Penetration tests for web apps, mobile apps, and LLM integrations

– Security audits with a focus on AI-powered systems and Agentic AI

– Security assessment of software architectures

– Mentoring for development teams on secure software development

– Consulting for companies integrating AI into their products

I prefer to work with medium-sized companies in the Freiburg region and throughout Germany. Personal contact is important to me – I am not an anonymous consulting firm, but an individual expert who is directly accessible.

My daily rate varies depending on the project scope and type. For an initial assessment, feel free to contact me – a brief initial consultation is always non-binding.