Work with us

Preamble: LLM agents (LLMA) are becoming the next-generation platform to develop intelligent systems. However, not much has been discussed on their energy consumption.​

Research questions: what are the sources of energy consumption in LLMA architectures? How to mitigate and control LLMA energy consumption? ​

What you will learn: what an LLM agent is, LLMA architectures, energy consumption reasoning, energy measurement, and technical skills ​

Collaborators: Henry Muccini @FrAmeLab, Rafiullah Omar @FrAmeLab, Prof. June Sallou @Wageningen University ​

#LLM_agent #sustainability​

Preamble: LLM agents (LLMA) are becoming the next-generation platform to develop intelligent systems. Much is ongoing in the field, with special attention to LLM agent architectures. However, guidelines on how to design agents are still missing.​

Research questions: Can we learn from microservice design best practices to engineering LLMA?​

​What you will learn: microservice design best practices, LLMA architectures, how to apply microservice best practices to LLMA​

Collaborators: Roberta Capuano @FrAmeLab, Rafiullah Omar @FrAmeLab, Henry Muccini @FrAmeLab 

#LLM_agents_design  #microservice architectures

Preamble: This research aims to explore how microservices can be designed and engineered with environmental sustainability in mind. Specifically, we investigate green design patterns and anti-patterns, empirical techniques for assessing the energy consumption of microservices, and both the state-of-the-art and industrial practices in this domain.​

Research questions: How are microservice architectures designed and engineered to support environmental sustainability in industry? What trends are emerging in green microservice practices? How do strategies differ across countries and regions?​

What you will learn: how to interview practitioners, how companies are greening their microservices, sustainable design​

Collaborators: Eoan O’Dea @FrAmeLab, Roberta Capuano @FrAmeLab, Henry Muccini @FrAmeLab, Prof. Dr. Verena Majuntke @HTW Berlin.​

 #Industrial_survey  #Green_microservices

Preamble: We would like to investigate how LLMs can be used to assist the architectural refactoring of software systems. In particular, ho to migrate from monolithic applications to greener microservice applications. For this purpose, we are investigating multiple LLMs and multiple prompt-engineering techniques. ​

Research Questions: What is the role of Large Language Models in enabling sustainability-driven decomposition of monolithic systems into microservices?​

What you will learn: How to apply LLMs and prompt engineering to identify microservice candidates, generate refactoring suggestions, and assess greener architectures using sustainability metrics.​

Collaborators: Roberta Capuano @FrAmeLab, Eoan O’Dea @FrAmeLab,  Henry Muccini @FrAmeLab,  Karthik Vaidhyanathan @IIIT Hyderabad

#LLM #Migration #Green_Microservices

Preamble: The role of Data centers in the consumption of ICT energy is mainstream. It is important to monitor and reduce their energy consumption. AI can play a role in making data centers more sustainable.​

Research Questions: How AI is being used for greening data centers? How AI could be used for greening data centers? Which are the best practices and challenges?​

What you will learn: how to run an empirical study, the role AI in greening Data centers​

Collaborators: Henry Muccini @FrAmeLab, Rafiullah Omar @FrAmeLab, Alex Montoya @TNO, The Netherlands

#Data_Centers #AI_4_sustainability  #Empirical_ Study

Preamble: This project focuses on mining open-source software repositories to identify and categorize microservices-based systems on potential energy inefficiencies. Large Language Models will be used to support the analysis by detecting code patterns, classifying inefficient practices, and interpreting project characteristics at scale. The resulting set of inefficiencies can help inform industry practices, supporting companies in addressing the increasingly important goal of building or refactoring applications for greener, more sustainable microservices.​

Research Questions: How can open-source software repositories be mined and analyzed to identify patterns of energy-inefficient code or practices? What characteristics of software projects correlate with higher potential source of energy consumption?​

What you will learn: Repository mining, static/dynamic analysis, code smells, energy inefficiencies, green software metrics, MSR methodology, empirical evaluation​

Collaborators: Eoan O’Dea @FrAmeLab, Roberta Capuano @FrAmeLab, Henry Muccini @FrAmeLab

#Green_Microservices #Mining_Software_Repositories 

Preamble: This project explores how Large Language Models can be used to assess whether two different implementations of the same system are functionally equivalent, even in the absence of an explicit requirements specification. By analysing the code and inferred behaviour, we aim to understand whether LLMs can help identify behavioural differences or confirm equivalence between versions.​

Research Questions: Can LLMs accurately compare two codebases to determine functional equivalence without relying on formal requirements? ​

What you will learn: How to design prompts and use LLMs to compare software implementations, infer intended behaviour, and assess whether different versions exhibit functional equivalence.​

Collaborators: Roberta Capuano @FrAmeLab, Eoan O’Dea @FrAmeLab,  Henry Muccini @FrAmeLab

 #LLM  #Functional_Equivalence #Code_Analysis

Preamble: As Extended Reality (XR) technologies become increasingly prevalent in industry—especially in sectors such as education, training, healthcare, and cultural heritage—there is a growing need to understand how software engineering (SE) practices are evolving to support gamified and immersive applications. Yet, little is known about how companies actually design, develop, and maintain these systems in real-world contexts. This project will investigate current SE practices through direct interviews with industry professionals working on XR solutions that include gamification elements.​

Research Questions: What are the most common software engineering practices used in the development of XR applications with gamification? How do teams approach requirements gathering, prototyping, and iterative design in such projects? What challenges do practitioners face in testing and maintaining XR systems? How do human factors and user experience considerations influence engineering decisions?​

What you will learn: How to design and conduct semi-structured interviews for qualitative research. Current software engineering practices in real-world XR and gamified system development. Challenges and trade-offs in requirements engineering, architecture, and testing in immersive systems. Qualitative data analysis using thematic coding techniques (e.g., grounded theory). How to synthesize findings into actionable research insights or design guidelines​

#Software Engineering #Extended Reality  #Gamification  #Industry Interviews, XR Systems #Requirements Engineering #System Design #SE Practices.

Preamble: Serious games in Cultural Heritage aim to enhance informal learning and user engagement. However, many current systems are static and fail to personalize the experience based on individual user behavior. This limits their educational potential. The proposed project focuses on designing an AI-based player modeling component to dynamically adapt gameplay, providing personalized, engaging, and educationally meaningful experiences for users exploring cultural heritage through mobile AR applications.​

Research Questions: How can player behavior be effectively modeled to support adaptive gameplay? Which AI techniques are most suitable for real-time user classification in cultural heritage games? How can adaptivity improve engagement and learning outcomes in serious games? What are the challenges of integrating AI components in mobile AR-based systems?​

What you will learn: How to design and implement adaptive systems in gamified and serious games environment. Techniques for player modeling and behavior analysis. Integration of AI components into modular architectures. Evaluation methods for adaptive educational systems.​

Collaborators: Martusciello Federico  @FrAmeLab, Rafiullah Omar @FrAmeLab, Henry Muccini @FrAmeLab

#AI #Adaptive Systems #Extended Reality #Gamification #Serious Games #Cultural Heritage​

​