tachAId: Interactive Tool Guiding Human-Centered AI Development

Pavlos Rath-Manakidis, M.Sc.
Researcher, Developer

Max Bauroth, M.Sc.
Researcher, Developer

Prof. Dr. Tobias Glasmachers
Supervisor

Prof. Dr. Laurenz Wiskott
Supervisor

Theory of Neural Systems

Funding:

Part of the HUMAINE competence center, funded by the German Federal Ministry of Education and Research (BMBF) within the Future of Value Creation-Research on Production, Services and Work program.

tachAId (Technical Assistance for Human-Centered AI Development) is an interactive guide that offers concrete recommendations for implementing human-centered Artificial Intelligence (AI). It is structured along the full AI lifecycle, from data collection and preprocessing to model design and deployment, with a particular focus on Automated Machine Learning (AutoML) and Explainable AI (XAI) to support effective human-AI collaboration. The tool aims to help various stakeholders—including AI engineers, future AI users, project owners, and unions—understand how to contribute to and benefit from human-centered AI, design AI for successful user collaboration, and understand what effective human-AI teamwork looks like.

Try tachAId

tachAId addresses key challenges in Human-Centered AI (HCAI), such as user interface design, user acceptance and engagement, human agency and oversight, and the critical aspects of fairness, transparency, and explainability. It provides context-specific guidance through an interactive "question-answer" format, tailoring recommendations to user needs. This approach bridges the gap between general human-centricity criteria and ethical principles, and the concrete technical steps needed for their implementation. tachAId also contextually recommends tools and libraries for XAI and AutoML. The tool is designed to be a process-accompanying resource, offering specific advice for various stages of technical implementation in AI/ML applications, particularly within a work context.

The development of tachAId is also a subject of a scientific publication, targeting researchers in HCAI, human-AI interaction, and work science. The publication discusses the interactive format, the underlying framework, and the methodology used in tachAId's creation. It also explores ethical principles, challenges, and goals for HCAI specific to workplace environments, and analyzes tachAId's scope in relation to the EU Ethics Guidelines for Trustworthy AI.

Link to publication:
Bauroth M, Rath-Manakidis P, Langholf V, Wiskott L and Glasmachers T (2024) tachAId—An interactive tool supporting the design of human-centered AI solutions. Front. Artif. Intell. 7:1354114. doi: 10.3389/frai.2024.1354114

Publications