Investigating Tacit Knowledge Transfer in Public Sector Workplaces
Iren Irbe · 2025
This poster paper presents the design and development process of TacitFlow, a voice-controlled assistant intended to support the transfer of tacit knowledge in high-stress public sector environments.Built using a Design-Based Research (DBR) approach, the system integrates voice input, knowledge graphs (KGs), Graph-Based Retrieval-Augmented Generation (GraphRAG), and graphof-thought (GoT) Artificial Intelligence (AI) reasoning for contextual support.The system enables knowledge capture, questionanswering, and peer learning through a mobile-first interface.This paper outlines the core functionality, architecture, and future evaluation plan.The design is grounded in findings from a scoping review and interviews, which are described in a separate Workin-Progress (WiP) paper titled: "Designing for the Unspoken: A Work-in-Progress on Tacit Knowledge Transfer in High-Stress Public Institutions" in ECCE 2025 proceedings.