Towards a Failure Analysis chatbot with Retrieval-Augmented Generation
Abstract
At Infineon, Failure Analysis (FA) is a fundamental business process that prevents and identifies causes of non-conformant behavior in semiconductor devices. The FA process is a complex and time-consuming task involving various specialized laboratory tests and research analysis. To boost chances of success, it is significant that FA engineers have quick access to relevant information for a specific investigation. Current FA information systems store data in various formats distributed across file shares and databases. Given the substantial amount of data, navigating and searching for particular data, analyzing its content, and merging the information is challenging. Simple keyword-matching interfaces are deemed insufficient due to the high workload of FA engineers. The emerging capabilities of Large Language Models (LLM) have been successfully applied in numerous applications such as summarization, question answering, translation, text classifica- tion, and chatbots. To enhance the efficiency of the FA research process and exploit the full potential of FA data, this thesis harnessed the power of LLMs in the domain of FA. A universal search interface was implemented as a chatbot to which users can pose interactive queries and receive straight-to-point answers. Since the native usage of LLMs underperforms in the downstream application of a complex data domain, Retrieval-Augmented Generation (RAG) was selected as a cost-effective adaptation technique. RAG enables the…
Read the paper
Related studies
- Knowledge Management Perspective of Generative Artificial Intelligence
- Navigating Generative Artificial Intelligence Promises and Perils for Knowledge and Creative Work
- Characterizing the evolving research on enterprise content management
- Large language models and multimodal foundation models for precision oncology
- From Generative AI to Extended Reality: Multidisciplinary Perspectives on the Challenges, Opportunities, and Future of Educational Computing
- Generative Artificial Intelligence: A New Engine for Advancing Environmental Science and Engineering
Go deeper in Evidence Hub
Evidence Hub adds a study data sheet (design, sample size, evidence level, key numbers), an AI summary and a chat grounded in the paper's text.