Given the data-intensive nature of Machine Learning (ML) systems in general, and Large Language Models (LLM) in particular, using them in cloud based environments can become a challenge due to legislation related to privacy and security of data. Taking such aspects into consideration implies porting the LLMs to an on-premise environment, where privacy and security can be controlled. In this paper, we study this porting process of a real-life application using ChatGPT, which runs in a public cloud, to an on-premise environment. The application being ported is AIPA, a system that leverages Large Language Models (LLMs) and sophisticated data analytics to enhance the assessment of procurement call bids. The main considerations in the porting process include transparency of open source models and cost of hardware, which are central design choices of the on-premise environment. In addition to presenting the porting process, we evaluate downsides and benefits associated with porting.

Containerization in multi-cloud environments has received significant attention in recent years both from academic research and industrial development perspectives. However, there exists no effort to systematically investigate the state of research on this topic. The aim of this research is to systematically identify and categorize the multiple aspects of containerization in multi-cloud environment. We conducted the Systematic Mapping Study (SMS) on the literature published between January 2013 and July 2024. One hundred twenty one studies were selected and the key results are: (1) Four leading themes on containerization in multi-cloud environment are identified: ‘Scalability and High Availability’, ‘Performance and Optimization’, ‘Security and Privacy’, and ‘Multi-Cloud Container Monitoring and Adaptation’. (2) Ninety-eight patterns and strategies for containerization in multi-cloud environment were classified across 10 subcategories and 4 categories. (3) Ten quality attributes considered were identified with 47 associated tactics. (4) Four catalogs consisting of challenges and solutions related to security, automation, deployment, and monitoring were introduced. The results of this SMS will assist researchers and practitioners in pursuing further studies on containerization in multi-cloud environment and developing specialized solutions for containerization applications in multi-cloud environment.

One distinctive characteristic of software is its malleability and succeeding opportunity and also need for a constant change. However, in certain types of software, many agencies in the public sector are bound by the laws on competition and financed by governmental budgeting processes. Consequently, the acquiring agencies adapt their processes to what the existing systems allow due to complexities in making changes and modifications. In this paper, we study the relation of public sector software acquisition and delivery, and the continuously changing nature of software. Then, we analyse pain points of evolving software in public sector information systems, and propose technical artefacts to improve the status quo. Finally, to add flexibility to public sector information systems’ acquisition and evolution, we elaborate a novel architectural pattern called Mosaic architecture, based on macroservices, which allows independent subcontracting, development, and deployment. The approach is demonstrated with a real-world industrial study from a public service that has been sketched applying the macroservice principles.