Work

Building Momentum in AI Implementation: An Action Research Study of Organisational Conditions

Year: 2026
Type: dissertation
Abstract: This Master’s thesis investigates how organisational conditions influence AI development and implementation in knowledge work. The research was conducted in a Nordic Private Equity company, where five AI tools were built as proof-of-concepts across four Action Research cycles, each structured by the Cyclical Process Model's stages of diagnosing, planning, building, evaluating and specifying learnings. This method of research was used to go beyond theory and see how AI development works in practice, when it has to compete as one against many other initiatives. The research revolved around bi-weekly meetings with senior organisation representatives, holding the organisation fixed, while teams and problems changed, allowing a direct comparison. Other evidence was collected via diagnostic interviews, a ranking form and per-cycle summaries. The conditions which the literature holds important were there, and some cycles did progress, but most tools never reached production. Three factors separated the two: how close the feedback providers were to the actual workflow, the depth of diagnosis related to the end-user's daily practice and how well the stated commitment led to contributed effort. The study contributes to the ongoing AI development research by extending the current literature through showing how established frameworks' reliance on binary conditions does not show the full picture, as in reality these conditions can behave more as degrees. In practice, the person whose daily work is impacted should be present as early as possible, with meaningful commitment allocated in the beginning, as well as throughout the project to judge the outputs and drive the development.
Author: Pham, Tien
Institution: Aalto University
Language: English
Open Access status: green
OpenAlex ID: W7204017312
Cites: 35
Cited by:
Related to: 10