Dr. Jan-David Stütz

Using my background in user-centered and neuro-symbolic AI in information systems from my PhD, I work on building advanced, semantically powered AI features for eccenca Corporate Memory - an enterprise knowledge-graph platform that connects and contextualizes complex data and business knowledge - to enable transparent and automation-ready knowledge management that drives smarter decision-making and scalable AI-driven workflows for clients, without black-box AI, the pain of siloed data, opaque semantics, or brittle automation.

Since transitioning to freelancing, I've focused on AI-powered product development at the intersection of engineering, design, and strategy. My work ranges from building full-stack web applications with modern frameworks (Next.js, React, Python, Flutter, Java, SpringBoot) to designing and implementing AI-based systems, including context engineering, knowledge graphs, and AI agents. I bring a holistic perspective to each project, shaped by my PhD in user-centered AI and years as a Senior Research Engineer at Bosch — combining stakeholder discovery, UX design, and full-stack implementation into a single delivery. My clients benefit from working with one person who can own the entire product lifecycle: from understanding the problem and designing the solution to shipping production-grade software.

At Bosch, I researched and developed user-centered and neuro-symbolic AI approaches for enterprise information systems. I designed methods to transform structured and unstructured data into usable knowledge graphs, enabling faster knowledge modeling, lowering the barrier for non-expert users to perform expert-level tasks, and significantly improving user experience and system transparency.

Building on these knowledge graph foundations, I developed purchasing-centric, knowledge-driven solutions that enhanced procurement processes by improving efficiency (time and cost savings), usability, and technology acceptance.

Additionally, I designed a goal-oriented recommendation approach in a DIY context that improved recommendation quality while mitigating common challenges such as cold-start problems. The system integrated knowledge-based reasoning with adaptive services, increasing efficiency and user satisfaction without sacrificing transparency or control.

The tech stack for my work included Python, TypeScript, Next.js, GraphDB, PostgreSQL, Azure, and Supabase. Additionally, my work included various modern AI approaches, such as RAGs and AI agents.