I'm starting my fourth year of Computer Science and need to choose an undergraduate thesis topic related to DevOps or cloud computing. I'd like to develop the thesis in collaboration with the company where I'm working, so they asked me to bring forward a few possible ideas before deciding on a project. I'm still relatively new to the field after beginning an internship in July, so I'm looking for something practical and achievable rather than an especially complex research topic. The company uses Microsoft Azure, Terraform, GitHub Actions, Docker, Linux, Python, Bash, and PowerShell. I've also gained some introductory experience with Azure monitoring, Azure Data Explorer and queries, cloud networking, and security. My work next year will involve fleet-related data. I'm hoping to study a process that is useful to the company while also becoming more familiar with the tools and workflows used there. What thesis subjects could fit this situation?
5 Answers
Kubernetes, Ansible, or more advanced deployment security are possible directions, but the right choice depends heavily on the company’s infrastructure and the access you can receive. In a large organization, the biggest challenge may be getting a sufficiently narrow scope and permission to use realistic data. Before settling on a topic, ask which team owns the relevant systems and what non-sensitive test environment is available.
A manageable DevSecOps topic would be adding tools such as Trivy and Gitleaks to the GitHub Actions pipelines. You could study how effectively they identify vulnerable container images or accidentally committed secrets, and then measure the effect on build duration and developer workflow. It’s practical, easy to prototype, and produces clear results.
That’s very actionable. Comparing the security benefits with the extra pipeline time could make the scope clear and realistic.
A good starting point is to look for a measurable improvement in an existing process rather than inventing a completely new system. Cloud cost analysis, deployment reliability, CI/CD performance, infrastructure monitoring, and security automation could all work well. Pick one concrete metric—such as money saved, build time reduced, deployment failures avoided, or vulnerabilities detected—and evaluate the current process against your proposed improvement.
Standardizing local development environments with Dev Containers could also make a strong project. You could compare the current setup process with a containerized version, measuring onboarding time, configuration errors, and consistency between developers. Since it builds on Docker and scripting without requiring a large production deployment, it may be suitable for someone still gaining experience.
Given the tools you listed, you could investigate automatically shutting down non-production Azure resources outside working hours and cleaning up unused disks or environments. The project would be fairly contained, and you could measure the monthly cost savings, implementation effort, and any operational risks. That gives the company a clear financial benefit while providing enough material for a thesis.
The company is quite large and already uses Kubernetes, although I haven’t encountered Ansible internally. I’m also trying to avoid exposing sensitive details, so narrowing the project to a specific team and test environment sounds useful.