Lars Lenssen Consultant as Freelancer Short Biography 10 years of project experience for energy utilities companies, e.g: E.ON Digital Technology and Wipro Limited. Dr.-Ing. from TU Dortmund University (March 2025), Artificial Intelligence Group under Erich Schubert. Wide experience in the german energy market, obtained under Uwe Strahmann. Proud father of two sons and passionate table tennis player. [LinkedIn] [XING] [Google Scholar] [freelancermap] [Email]

Lenssen Consulting Deep expertise in enterprise systems, data-driven development, and project delivery. Project leadership, from managing cross-functional teams to coordinating offshore delivery models. Deep scientific and theoretical expertise in AI, which I actively apply to help organizations transform their processes and structures through practical, impactful solutions.

Selected Projects

• Cloud Migration Architecture for Multi-Stack Applications Duration: Since 2025 (ongoing) Context: Azure Cloud Services Description: Cloud migration of applications across diverse technology stacks. Modernized legacy systems to ensure cloud compatibility, enabling them to run efficiently in a scalable and maintainable cloud environment. Focused on integrating modern infrastructure, reducing technical debt, and enhancing long-term flexibility and performance.

• DSGVO-Compliant Data Lifecycle Automation Duration: Since 2024 (ongoing) Context: SAP IS-U, SAP ILM Description: Developed and implemented a high-performance solution for archiving and deleting large volumes of legacy enterprise data. Focused on automation to reduce manual workload and ensure consistent, error-free execution. The solution was designed to support DSGVO compliance and improve overall system performance and maintainability.

• Implementation of the Energy Price Brake in Utility Billing Systems Duration: 11 Months (2023) Context: SAP IS-U: Common Layer Description: Designed and implemented the integration of the energy price brake into existing billing systems. Translated regulatory requirements into technical solutions to ensure accurate and compliant billing processes. Delivered a reliable and maintainable implementation within the SAP IS-U environment, supporting energy providers in adapting to new legal frameworks.

• Implementation of Market Communication 2022 (MAKO2022) Duration: 10 Months (2022 - 2023) Context: SAP IS-U: Common Layer, IDXGC, IDXGL Description: Conceived and implemented the requirements for MAKO 2022, focusing on the integration and handling of communication formats such as MSCONS, UTILMD, and PARTIN. Ensured compliance with regulatory standards and seamless data exchange within the energy market communication framework.

• Team Leadership and Offshore Delivery Design Duration: 12 Months (2020 - 2021) Context: Applications on different technology stacks: Jboss, Sitecore, EAI, Selenium Testautomation, Azure Cloud Services, SAP IS-U. Description: Led a support and development team of 15–20 members across multiple technology stacks. Responsible for coordinating with internal and external stakeholders, managing day-to-day operations, and designing an effective offshore delivery model to improve scalability and cost-efficiency. Ensured seamless collaboration between onshore and offshore teams while maintaining high service quality.

• Support Team Coordination and Requirements Development Duration: 11 Months (2018 - 2019) Context: SAP IS-U / CRM Description: Managed a support team of 8–10 members, ensuring efficient incident handling and service continuity. Acted as the main liaison between technical teams and stakeholders, translating business needs into technical requirements. Led the design and implementation of new features and enhancements to meet evolving client demands.

Publications

• Archetype Discovery from Taxonomies: A Method to Cluster Small Datasets of Categorical Data Proceedings of the 58th Hawaii International Conference on System Sciences Lars Lenssen, Philip Stahmann, Christian Janiesch and Erich Schubert Short Abstract: This work addresses this gap by exploring information-theoretic approaches to develop a novel clustering method CatRED tailored for small categorical datasets such as taxonomy data. We evaluate our method through its application to two taxonomy datasets, demonstrating its effectiveness in generating archetypes. [Paper] [Code]

• Medoid Silhouette clustering with automatic cluster number selection Information Systems 120, 102290, 2024 Lars Lenssen and Erich Schubert Short Abstract: There are many different clustering quality measures, to validate clustering results. We discuss the efficient medoid-based variant of the Silhouette, and provide two fast versions for the direct optimization. Additionally, we provide a variant to choose the optimal number of clusters directly. [Paper] [Code]

• Sparse Partitioning Around Medoids Machine Learning under Resource Constraints -- Fundamentals 1, 182-196, 2023. Lars Lenssen and Erich Schubert Short Abstract: Partitioning Around Medoids (PAM, k-Medoids) is a popular clustering technique to use with arbitrary distance functions or similarities. By exploiting sparsity, we can avoid the quadratic runtime and memory requirements, and make this method scalable to even larger problems. [Paper]

• Fast k-Nearest-Neighbor-Consistent Clustering LWDA 2023 (KDML: Best Paper Award) Lars Lenssen, Niklas Strahmann and Erich Schubert Short Abstract: We propose a fast variant of the K-means clustering algorithm that uses the k-Nearest-Neighbor Consistency as a constraint. [Paper] [Code]

• Clustering by Direct Optimization of the Medoid Silhouette SISAP 2022 (Best Student Paper Award) Lars Lenssen and Erich Schubert Short Abstract: There are many different clustering quality measures, to validate clustering results. We discuss the efficient medoid-based variant of the Silhouette, and provide two fast versions for the direct optimization. [Paper] [Code]

• Fast k-medoids Clustering in Rust and Python Journal Open Source Software 7(75), 4183, 2022. Erich Schubert and Lars Lenssen Short Abstract: We introduce the kmedoids Rust crate along with a Python wrapper package kmedoids to make this fast algorithm easier to employ by researchers in various fields. [Paper] [Python] [Rust]