DBIS

Research Questions

This thesis will address the following core scientific and engineering questions:

• Design and Implementation of AI Agent Systems
  • What core capabilities are required for AI Agents to understand user intentions, execute tasks autonomously, and coordinate cross-departmental collaboration?
  • How can AI be leveraged to automate complex enterprise processes such as approvals, project management, and interdepartmental workflows?
  • Within the context of small and medium-sized enterprises (SMEs), how can customization and flexibility be balanced with the need for standardized and scalable automation?
• System Integration and Cost Efficiency
  • How can seamless integration between AI Agents and existing enterprise software (e.g., ERP, CRM, OA systems) be achieved while maintaining cost efficiency?
  • To what extent can cloud-based deployment, open-source solutions, or modular/plugin-based architectures lower the adoption barriers for SMEs?
• Human–Computer Interaction and User Experience
  • How can natural language interfaces be designed to enable AI Agents to better understand enterprise users’ tasks and organizational context?
  • How can efficiency gains be achieved while ensuring that users maintain a sense of control and trust in intelligent agents?
• Compliance and Data Security
  • In the German enterprise environment, how can AI Agents ensure that data processing complies with GDPR and other relevant regulations?
  • How can intelligent decision-making be reconciled with organizational requirements for transparency and explainability?

Objectives and Tasks

To achieve these objectives, the research will be structured into the following tasks:

1. Enterprise Software Prototype Development

• Build a lightweight enterprise software prototype with core components including instant messaging, calendar, email, documents and wiki (knowledge base), project and task management, and meetings collaboration.
• Implement basic user management and access control to simulate common SME operation and project management scenarios.

2. AI Agent System Design and Integration

• Design AI Agents with capabilities for intent recognition, task automation, and workflow coordination across different enterprise functions.
• Integrate these agents into the prototype system to enable proactive assistance in tasks such as approvals, scheduling, project tracking, and document management
• Investigate methods for seamless integration between the AI-driven system and existing enterprise software (e.g., ERP, CRM) while minimizing deployment and maintenance costs.
• Explore cloud-based, open-source, and plugin-based architectures to enhance scalability and reduce adoption barriers for SMEs.

3. Compliance and Security 

• Propose auditability, explainability, and data protection mechanisms for AI Agents in line with GDPR and other regulatory requirements.
• Integrate logging and access isolation into the prototype system to ensure transparency, controllability, and compliance of AI Agent actions.

4. Cost and Scalability Optimization

• Explore lightweight deployment models (e.g., hybrid cloud and on-premises) to reduce implementation and maintenance costs for SMEs.
• Research modular and plug-and-play designs, enabling AI Agents to be gradually integrated into enterprise software across different industries and scales.

5. System Evaluation

• Conduct case studies and experiments to assess the performance of AI Agents in terms of workflow automation, project execution efficiency, user satisfaction, and compliance assurance.
• Collect user feedback to evaluate the feasibility of AI Agents in human–machine collaboration and project management acceptance.