We offer innovative and impactful topics at the intersection of Artificial Intelligence and Management. In particular, we favour topics related to organisational, social and environmental implications of AI. We jointly select topics based on interest, fit and the study focus of the students. We are also happy to supervise thesis based on industry internships that have elements of AI or emerging technologies in it (but the internships must be organised by students themselves).
Ready to apply? Fill in this form –> Master Thesis Supervision
Have questions? Reach out to Henri –> henri.jamet@unil.ch
In our selection procedure for MSc Thesis supervision, we will prioritise students who have participated in one of our courses and successfully passed the exams. Please find below a list of example thesis topics at our lab.
Open Thesis Topics
1. Generative AI for Business
Recent advancements in computational power and data availability have enabled the development of increasingly complex forms of artificial intelligence (AI) systems, particularly those that rely on machine learning and deep neural network (DNN) architectures (Shrestha et al., 2021). One notable recent development in this field are generative AI (GAI) models (also known as foundation models or large language models (LLMs)), such as BERT and GPT-3. These models utilize specialized DNN architectures that are trained on a vast quantity of unlabeled data, such as large fractions of the internet, at scale and in parallel, using hundreds of computers (Bommasani et al., 2021; Longoni et al., 2022). These developments have led to a rapid proliferation of applications, such as Stable Diffusion, DALL-E2, Flamingo, Florence, and ChatGPT, which are currently transforming the AI landscape. Given their ability to generate original human-like textual, visual, and auditory content with little or no human input and intervention, drawing entirely on the data on which they have been trained, these technologies are taking assistive technology to a whole new level. They enable users to incorporate generated outputs into their creative work, reducing application development time and bringing powerful capabilities to even non-technical users (Chui, Roberts & Yee, 2022). For instance, ChatGPT has recently been used to produce poetry, advertising blogs, and sketch package designs, as well as to write computer code and identify bugs in existing code. GPT-3 is being utilized by scientists to create novel protein sequences and review scientific literature. These examples demonstrate the potential of GAI to significantly impact various fields and industries.
GAI systems are considered a turning point in the field of AI and are widely believed to have the potential to revolutionize a wide range of industries and applications, ranging from marketing and sales, operations, engineering, human resource management, risk and legal to research and development (Mollick, 2022). For instance, GAI can assist in marketing and sales by creating user guides, analyze customer feedback, identify potential risks, recommend next steps for marketing and sales interactions, and improve sales support chatbots (Chui, Roberts & Yee, 2022). In human resource management, GAI can be used to create interview questions for candidate assessment and automate first-line interactions such as employee onboarding. The advent of GAI is currently driving the rapid industrialization of AI, as companies are adopting and customizing existing foundation models into their business processes and products. Anticipating the future of GAI, Microsoft has reportedly planned to invest $10B in ChatGPT’s creator company OpenAI[1].
Despite the excitement surrounding the potential of GAI to enhance creativity and innovation and deliver unprecedented opportunities, companies are also facing organizational, social, and environmental concerns as they integrate GAI into their operations (Bailey et al., 2022; Raisch & Krakowski, 2021; Shrestha et al., 2019, 2021). First, a clear challenge is the difficulties inherent in collaboration and coordination between human and GAI (Feuerriegel et al., 2022). These concerns extend to the use of GAI detrimentally impacting individual learning and creativity. Second, GAI application by organizations also raises social concerns, such as the potential for algorithmic biases, misuse of technology for large-scale and automated fakery, and intellectual property (IP) risks related to commercialization of products developed using GAI inputs. Third, development of GAI requires significant energy consumption, which raises environmental challenges. To fully leverage the benefits of GAI, it is crucial for organizations to address these challenges by developing solutions and frameworks for effectively integrating GAI into their operations (Feuerriegel et al., 2022). This research project aims to investigate some of these challenges and identify potential solutions to overcome them.
[1] https://www.cnbc.com/2023/01/10/microsoft-to-invest-10-billion-in-chatgpt-creator-openai-report-says.html
2. Fairness in Recommendation Systems (MSc) with Stanford University
Goal of this project is to evaluate currently available food recommender systems and build a recommender system that takes into consideration healthy food as a recommendation set. Knowledge of Computer Vision (image recognition) and recommender system is useful for this project.
3. Deep Learning for Disease Diagnosis (MSc) with Stanford University
Augmentation of disease diagnosis and decision-making in health care with machine learning algorithms is gaining much impetus in recent years. In particular, in the current epidemiological situation caused by COVID-19 pandemic, swift and accurate prediction of disease diagnosis with machine learning algorithms could facilitate identification and care of vulnerable clusters of population, such as those having multi- morbidity conditions. In order to build a useful disease diagnosis prediction system, advancement in both data representation and development of machine learning architectures are imperative. In this project we seek to build deep learning pipeline for efficient and accurate diagnosis of diseases.
Literature: See here
4. Deep Learning on Text Documents and Knowledge Graphs (MSc) with Stanford University
Deep learning methods have given rise to a variety of models to learn representations (or embeddings) from both structured and unstructured data. In this project we will evaluate some existing methods and develop new techniques to learn embeddings from text documents and knowledge graphs. Especifically, the goal is to extract information from text documents, so that they can be populated into knowledge bases. Working on this project requires experience with both structured and unstructured datasets, and of the mathematical models to represent them. Coursework on NLP, ML, or Statistics will be helpful. Experience with Tensorflow and/or Keras, or the readiness to learn them will be expected.
Relevant literature:
– Weston, Jason, et al. “Connecting language and knowledge bases with embedding models for relation extraction.” (2013).
– Zhang, Wen, et al. “Interaction Embeddings for Prediction and Explanation in Knowledge Graphs.” (2019).
See also here
For more information or if you want to apply for the thesis, please contact me or Bibek Paudel (bibekp@stanford.edu)
5. Organizations and Algorithms
During the last two decades, the scientific advancements in artificial intelligence (AI) and ML algorithms in the field of computer science—including findings from research on deep neural networks and the development of hardware such as graphics processing units (GPUs) and tensor processing units ( TPUs) that makes training such complex nonlinear models feasible—has led to the development of predictive technologies capable of undertaking tasks that previously required human judgment and decision making. Such, AI- and ML-based predictive technologies have been successfully applied to automate knowledge work and decision making (eg, in dynamic pricing in e-commerce websites and high-frequency trading and recommender systems). When well integrated, firms can benefit from partially automated/fully automated decision making because (in certain scenarios) it reduces coordination costs and frees up human attention. However, given the opacity, lack of accountability, and embedded bias of these algorithms, managers might lack the required expertise to better leverage predictive technologies in their organizations. Given their idiosyncratic features, there is a need to rethink organizational designs when integrating these predictive technologies into our organizations. This gives rise to the grand challenge, namely, how to redesign organizations to efficiently benefit from the automation of knowledge work and decision making. The focus of work in this thesis should aim to understand how organizations and organization designs are being shaped by algorithms as both tools and agents. Working on this project requires experience with both structured and unstructured datasets, and of the mathematical models to represent them. Coursework on NLP, ML, or Statistics and those on strategy and organization theories will be helpful.
Relevant literature:
– von Krogh, G. (2018). Artificial Intelligence in Organizations: New Opportunities for Phenomenon-Based Theorizing. Academy of Management Discoveries, 4(4), 404-409.
– Shrestha, YR, Ben-Menahem, S., & von Krogh, G (2019). Organizational Decision-Making Structures in the Age of Artificial Intelligence. California Management Review.
Past Master Thesis Supervision @ UNIL
| Year of Graduation | MSc Student | Prof. | Main Tutors | Company | Topic |
| 2024 | Ahmad Farhat | Yash | Yuanjun Feng and Gregory Chevalley | Stouff Capital | Adapting to Evolving Data Needs:Integrating No-Code Visualization in a Private Investment Firm – Stouff Capital Case Study |
| 2024 | Mireille Adjaoute | Yash | Yuanjun Feng | ABBL | Data Complexity in Open Finance Integration Challenges and Solution Alternatives |
| 2024 | Srivathshan Paramalingam | Yash | Amirsiavosh Bashardoust and Arnaud Rambicur | KPMG | Optimizing Artificial Intelligence Integration in Auditting: Enhancing Quality and Addressing Challenges |
| 2024 | Giacomo Rattazzi | Yash | Amirsiavosh Bashardoust and Henri Jamet | Academic | Analyzing the Political Bias of Large Language Models on Climate Change Discourse |
| 2024 | Yonah Bole | Yash | Amirsiavosh Bashardoust | Ubisoft | Workflow Efficiency in Game Development: Exploring How Integrating Generative Artificial Intelligence, Specifically Stable Diffusion, Into Game Development Pipelines Can Enhance Efficiency, Creativity, and Innovation. |
| 2024 | Ismael Tewfik | Yash | Yuanjun Feng | Richemont | Optimizing Onboarding with “Enboarder”: A Solution to Early Employee Turnover |
| 2024 | Krzysztof Szerenos | Yash | Yuanjun Feng | Santander Private Bank | Optimizing Bank Operations: A Case Study on the Appian Case Management Tool Implementation in Santander Private Bank |
| 2024 | Christophe Rosset | Yash | Henri Jamet | Vaudoise Assurances | How DevOps Tools Can Enhance Software Engineering Activities for HR Software Deliveries ? |
| 2024 | Johannes-Rudolf David | Yash | Henri Jamet | Visible SA | LIFT: Leveraging LLMs to Assist Meeting Facilitators |
Past Master Thesis Supervision @ ETH Zurich
19. Claudio Zihlmann (2022)
Topic: Market access strategy for a new value product in the oral market for biomaterials MAS Excellence Award 2022
18. Lydia Pagani (2022)
Topic: Stylistic production: How the stylistic choices of the occupational community of the Swiss independent watchmakers influence their company’s strategy
17. Savindu Herath (2021)
Topic: Leaveraging data-driven decision-making in co-creation business models to improve firm performance: Evidence from online fashion retailing Received ETH Medal Award 2022 (top 2%)
16. Tobias Motz (2021)
Topic: Artificial Intelligence and Organizations: Paradigms of action for a successful integration
15. Anastasios Papageorgiou (2021)
Topic: Deep Learning for disease diagnosis
14. Leopold Franz (2020)
Topic: Managing Disease Diagnoses with Structured and Unstructured Clinical Data\
13. Sebastian Windeck Otto (2020)
Topic: Knowledge graphs for strategic business applications
12. Martin Buttenschön (2018)
Topic: Data for AI: How well structured data empowers business to benefit from machine learning
11. Apurva Maduskar (2017)
Topic: When Corporate Agile Meets Open Source: Contrasting Knowledge Integration and Documentation Practices Recieved MAS Excellence Award 2018
10. Chun-Hui Kuo (2017)
Topic: From startup to scaleup: Two-phase searching for human resource acquisition inearly-stage spin-offs
9. Matthias Stenske (2017)
Topic: Open Source Strategy for Swiss Telecommunication infrastructure industry:Impact on strategic resources
8. Mikko Leimio (2017)
Topic: The Impact of an Open Source Hardware Strategy for 5G Technology to the Telecom Industry and a major Swiss Telecommunications Provider
7. Matteo Frondoni (2016)
Topic: Competing with giants: Artificial Intelligence as a threat or an opportunity for the swiss TIME industry?
Recieved Student Prize 2016 from the SEW-EURODRIVE Foundation and ETH Medal 2017 (top 2%)
6. David Roegiers (2016)
Topic: Start-up acquisitions in the open source software space: What is the effect oncommunity dynamics ?
5. Christoph Hirnschall (2016)
Topic: Online learning of user preferences with applications for online marketplaces
Recieved Willie Studer Prize 2017 (best student in each ETH Zurich Master’s degree programme)
4. Matthias Auf der Mauer (2015)
Topic: Business Model Analysis of Intuitive Surgical Inc.and Strategic Implications for Companies in the Robotic Assisted Ablation Catheter Industry
3. Pascal Mages (2015)
Topic: IT Outsourcing for Small and Medium Enterprises
2. Remo Hug (2015)
Topic: Sharing economy in the market for kid’s goods in Switzerland
1. Fotini Traka (2015)
Topic: OSS Licenses and project sustainability