Modular AI Framework for Small Software Companies
Niraula, Saroj (2025)
2025
All rights reserved. This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:amk-2025121235272
https://urn.fi/URN:NBN:fi:amk-2025121235272
Tiivistelmä
This thesis presents the conceptual design of a Modular AI Framework intended to help small and medium-sized software companies adopt artificial intelligence in a structured and sustainable way. The study applies a design science research approach to identify adoption barriers, define framework requirements, and propose a three-layer architecture consisting of Module, Integration, and Management layers. Each layer supports modularity, interoperability, and maintainability while remaining lightweight enough for small development teams.
The framework was conceptually evaluated through illustrative case scenarios, such as a natural language processing service and a model management module. These examples demonstrated how containerization, standardized APIs, and clear module specifications can reduce integration effort and long-term maintenance costs.
The results suggest that modularity enables SMEs to adopt AI incrementally, reuse components across projects, and avoid vendor lock-in. Although no full implementation was developed, the conceptual evaluation confirms the framework’s potential to improve accessibility and sustainability of AI solutions for smaller organizations. The study concludes that modular architecture provides a pathway for SMEs to participate in AI innovation while maintaining technical and regulatory flexibility.
The framework was conceptually evaluated through illustrative case scenarios, such as a natural language processing service and a model management module. These examples demonstrated how containerization, standardized APIs, and clear module specifications can reduce integration effort and long-term maintenance costs.
The results suggest that modularity enables SMEs to adopt AI incrementally, reuse components across projects, and avoid vendor lock-in. Although no full implementation was developed, the conceptual evaluation confirms the framework’s potential to improve accessibility and sustainability of AI solutions for smaller organizations. The study concludes that modular architecture provides a pathway for SMEs to participate in AI innovation while maintaining technical and regulatory flexibility.