Through my extensive project experience in applied artificial intelligence, with a strong focus on generative AI, dialogue systems, and hybrid intelligence, I have developed comprehensive expertise in designing and implementing trustworthy AI solutions for organisational and societal contexts. My work combines methods from machine learning, knowledge engineering, and digital business process management. I typically follow an agile project management approach in combination with design science research, ensuring that innovative AI concepts are not only theoretically grounded but also validated in practice.
In the RepoChat project, a dialogue system with a large language model / LLM, retrieval mechanism and knowledge graph is to be researched and implemented, which should enable the addressee-oriented accessibility of Nagra reports via a chat interface and make statements verifiable.
The aim of the research project is to demonstrate the positive health effect of the use of an intelligent health chatbot by young Nigerian adults living with HIV. In addition to the FHNW, the Faculty of Public Health Department of Health University of Ibadan and the Department of Epidemiology and Public Health Swiss Tropical and Public Health Institute are project partners.
The project elaborates an approach that integrates long-context large language models (LLMs), human expertise, and hybrid clustering mechanisms to detect and analyse inconsistencies in large-scale document repositories.
In the ChEdventure project, a simulation environment is being developed in which students can practise complex issues through dialogues that often occur in business workshops. ChEdventure offers virtual interlocutors with whom students can learn skills such as asking relevant questions, critically questioning answers, placing them in the overall context and resolving contradictory statements.
Design of a new degree programme for the School of Business FHNW in the field of artificial intelligence, business and socio-technical information systems.
The purpose of this Innosuisse’s innovation cheque project is to test the feasibility of the idea to offer a platform that enables organisations such as law firms or associations to develop their own chatbots for legal advice ("legal chatbots") with little effort. The project partners are the Velex GmbH, the FHNW Intelligent Information Systems Research Group lead by Prof. Dr. Knut Hinkelmann and FHNW Competence Center Cloud Computing, Digitalisation & Transformation lead by Prof. Dr. Stella Gatziu Grivas.
First, the cognitive adequacy of business process models was described and assessed. Based on these findings, a business process modelling language was selected to describe process fragments. Finally, a new combined approach consisting of process execution and case-based reasoning was described, evaluated and instantiated as a software demonstrator called ICEBERG-PE. The thesis contains a well-founded literature framework and follows a design science research (DSR) strategy. Four conference and workshop publications, as well as a journal article and one book chapter, have emerged from this work. The triangulated evaluation of the ICEBERG-PE instantiation confirmed the transferability of this approach.
The aim of the project was the implementation of knowledge management supporting software integration projects using case-based reasoning and an enterprise ontology. The project goals were the development of an ontology-based context model, the elaboration of a structured language for describing integration project knowledge and the prototypical implementation of an ontology-based reasoner.
The Eurostars project LeadAI develops innovative AI-driven tools to support leadership consulting through organisational digital twins. Within this project, the focus lies on agentic AI and LxM agents, particularly in the development of a Virtual Personal Assistant prototype. Contributions include the design of orchestration and data protection components, as well as leadership assessments such as tailored questionnaires and structured repositories of expert inputs.
The aim of this Innosuisse Innovation Check project was to test the feasibility of the idea of a digital innovation assistant. The project partners are the Superloop Innovation GmbH and FHNW Competence Center Cloud Computing, Digitalisation & Transformation under the lead of Prof. Dr. Stella Gatziu Grivas.
This didactic development project investigated how a prototypical chatbot as an intelligent and dialogue-oriented tutoring system supports coaching in project-related teaching and thus supports the respective learning progress of individual students working in teams.
The aim of the project was to develop a digital assistant, which supports the customers of a BI consulting provider in the selection and configuration of business intelligence services in the cloud. The assistant is based on a case-based reasoning-based recommender that accesses a case base and automatically derives recommendations to the customers.
More complex requirements for SMEs force the transition from a supplier- to a network-management to handle hidden risks adequately. With the implementation of a case-based reasoning approach, tools for “Supply Network Management” have been made available.
In the 21st century, progressive digitalisation has brought forth new types of machines and tools and made new business models and forms possible. This project provided guidance on how the education and training of the FHNW School of Business should adapt to progressive digitalisation.
Learn PAd was an interdisciplinary project with the aim of providing learning solutions for public administrations that focus on their business processes and context.
The scientific goal of the project was to transfer semantic technologies into the application. By using formalised contextual information and classical information extraction, the content of a document is to be made accessible and metadata automatically generated.
The EU project MATURE was based on the concept of knowledge maturation, which is understood as goal-oriented learning on a collective level. The project explores how knowledge maturation takes place within and between organisations, what barriers exist and how socio-technical solutions can be used to overcome/manage these barriers.
Management consultancy in the areas of digitalisation, business processes, artificial intelligence, organisational development and strategy management, as well as innovation and project management. Consulting mandates include advice to government bodies, municipalities, and school administrations, and the creation of ICT concepts and tender management.
Consultant for Business Process, Case Management and Decision Modelling, Conversational AI, Workflow Automation, Enterprise Integration and OMG Specs (BPMN, CMMN, DMN).
The project is developing an AI Tutor to assist students in learning and to research the impact of AI on learning processes.
- In mandate of the School of Business FHNW
- Contribution of the Conversational AI expertise
In this consulting project, a prototype was developed using large language models (LLMs) to analyse contracts and identify high-risk clauses and entities. The solution demonstrates how generative AI can support legal practitioners in compliance, risk management, and contract review.