Background

In today's fast-paced innovation landscape, understanding competitor behavior and emerging technology trends is crucial for strategic decision-making. Patent databases contain a wealth of information about technological developments, but manually analyzing thousands of patent documents is time-consuming and often impractical. With recent advances in Large Language Models (LLMs), there is now an opportunity to automate and enhance patent analysis at scale.

This thesis is conducted in collaboration with Trelleborg, a global engineering group specializing in polymer technology and industrial solutions. Trelleborg seeks to leverage AI-powered patent screening to gain competitive intelligence, identify emerging product developments, and detect technology trends within their industry. By applying state-of-the-art language models to patent data, this work has the potential to transform how organizations monitor the innovation landscape and make strategic R&D decisions.

The challenge lies in developing robust methods to extract meaningful insights from complex patent documents, which often contain technical jargon, legal language, and highly specialized content. Key questions include: How can LLMs effectively identify relevant patents among thousands of documents? What techniques enable the extraction of competitor strategies and technology trends? How can these insights be structured and communicated to support business decision-making?

In this thesis you will:

• Extract and structure insights on competitor activities, product developments, and technology trends from patent data
• Evaluate your approach using real-world patent data in collaboration with Trelleborg
• Deliver actionable insights and recommendations for strategic decision-making at Trelleborg

Research Impact

Your work will contribute to:

• Competitive intelligence: Enable organizations to systematically monitor competitor patent activities
• Trend detection: Identify emerging technologies and innovation patterns early
• Strategic R&D: Support data-driven decisions in product development and technology investments
• Scalable patent analysis: Demonstrate how AI can transform patent screening from manual review to automated intelligence

We are looking for candidates who:

• Have a strong interest in natural language processing and practical AI applications
• Are motivated to work on real-world business challenges with industry partners
• Show initiative and can work independently while collaborating with stakeholders
• Have strong analytical and communication skills (English and German required)

Details

• Start: Immediately
• Language: English and German
• Industry Partner: Trelleborg

How to Apply

We offer you an exciting research topic at the intersection of AI and business intelligence, close supervision and mentorship, direct collaboration with an industry leader, and the opportunity to work on a problem with real-world impact.

Interested in shaping the future of patent intelligence?

Please send your current transcript of records, a short CV, and a brief motivation (3-4 sentences) to daniel ∂does-not-exist.hendriks kit edu.