Thesis: Software Security Micro-Trainings using Large Language Models

Welcome to the Fraunhofer Institute for Mechatronic Design IEM!
At the »Zukunftsmeile« in Paderborn, Germany, we conduct practical research to develop innovative solutions for mechanical and plant engineering, the automotive industry and related sectors. The focus is on intelligent products, production systems, services and software applications.

What you will do

Given the barriers to traditional training methods such as insufficient time and resources, other industries have explored alternative learning models such as micro-learning. In this learning model, the content is broken into smaller pieces so that learners can quickly learn and apply new skills in real-world situations using interactive elements and gamification.
 

Software security trainings could possibly benefit from such a training approach by supporting tailored learning experiences based on the needs and schedules of the software developers. With recent advances in Large Language Models (LLMs) and their use for code generation, completion, explanation, refactoring and other software engineering tasks, exploring their application for also training software developers is worth further investigation. Pairing the capabilities of LLMs in understanding and generating natural language with software security training materials could possibly provide micro-training solutions for developers within their Integration Development Environment (IDE).
 

Such an approach would be useful in providing software developers with the direct support that they need, when a security vulnerability is detected in code that they push or also while developing new software components that might be susceptible to certain vulnerability vectors. Developers would be supported both in the preventative and post vulnerability detection phases so that new skills can be applied in real-world situations quicker, therefore increasing the learning outcomes.

In this thesis, you will work on the following tasks: 

• Conduct a literature review of publications and current trends in the domain of LLMs for trainings
• Develop and evaluate a prototype that uses LLMs to create dynamic software security micro trainings

What you bring to the table

• You are studying Computer Science or a comparable course of study  
• Good Java and/or Python programming skills  
• Machine learning knowledge and experience
• Experience with Static Analysis is recommended
• Good language skills in German and/or English

What you can expect

• A strong team culture with flat hierarchies is a matter of course for us. This means: high esteem and trust
• Professional supervision and specialist support in the preparation of the student research project/thesis
• Insight into the current challenges of the company
• Flexible working from home to combine studies and work in the best possible way

We value and promote the diversity of our employees' skills and therefore welcome all applications - regardless of age, gender, nationality, ethnic and social origin, religion, ideology, disability, sexual orientation and identity. Severely disabled persons are given preference in the event of equal suitability.

With its focus on developing key technologies that are vital for the future and enabling the commercial utilization of this work by business and industry, Fraunhofer plays a central role in the innovation process. As a pioneer and catalyst for groundbreaking developments and scientific excellence, Fraunhofer helps shape society now and in the future.
 

Interested? Apply online now. We look forward to getting to know you!

Additional questions will be answered gladly by:
Mr. Oshando Johnson and Ms. Samira Taaibi
Fraunhofer-Institut für Entwurfstechnik Mechatronik IEM 
Zukunftsmeile 1 | 33102 Paderborn 

Web: https://www.iem.fraunhofer.de/

Requisition Number: 78426