Successfully completed Bachelor Thesis by our working student Tim
NovelSense is now equipped with a scientifically grounded and practically validated methodology to successfully implement Agentic AI in German SMEs. By integrating the latest scientific findings with real-world application, we can now guide companies precisely from the first strategic idea to a specified technical use case. This extended capability is the direct result of the Bachelor’s thesis successfully completed by our working student, Tim Tasol.
Over the past six months, Tim dedicated himself to solving one of the most pressing challenges in the current tech landscape. While the hype around autonomous agents is huge, reliable empirical data is often scarce, and companies frequently struggle to bridge the gap between a strategic vision and a concrete technical implementation. To address this, Tim developed a specific procedural model designed to guide us and our clients through this complexity efficiently.
Deep dive into processes: Tim Tasol explaining an Agentic Workflow during the Kick-off at Neuland GmbH.
Source: Marc Schulze
The result is a robust practice guide derived from current state-of-the-art literature, in-depth interviews with over ten industry experts, and real-world validation through a pilot project with our client Neuland GmbH.
A key insight from the research is that Agentic AI acts more like a digital employee than a passive IT tool. This requires a much higher level of process mapping and governance, meaning compliance and risk management must be integrated from the very beginning.
Crucially, the developed model includes a strategic focus on finding the best solution for the client. While the framework is specialized for Agentic AI, its systematic approach ensures that if a standard automation or generative AI solution is more effective for the specific problem, that path is chosen instead.
The developed framework: A structured path to identifying and specifying Agentic AI use cases.
For our clients, this means we now have an expert-validated framework to identify high-value AI use cases securely. We would like to thank Neuland GmbH for their openness and collaboration in validating this approach. And finally: We are thrilled with the depth of expertise built during this project. Congratulations, Tim, on this successful degree!
Looking to evaluate Agentic AI use cases safely and pragmatically? Feel free to reach out.