Nicolai Benz

Title: Nicolai Benz and Enterprise RAG Evaluation

Introduction

Nicolai Benz is an inventor based in Mountain View, California, and works at SAP SE. His patent work addresses the evaluation and optimization of enterprise retrieval-augmented generation (RAG) systems.

Latest Patents

Benz is associated with 1 patent, titled “Enterprise Retrieval-Augmented Generation Evaluation Framework.” The invention describes a framework that evaluates an enterprise RAG system connected to one or more large language models. Using a test dataset containing questions, context information, and ground-truth answers, the framework measures generation, retrieval, and end-to-end performance. It then automatically adjusts at least one RAG tuning parameter and repeats the evaluation until a specified threshold is met. Once that threshold is satisfied, the framework can produce a report of the evaluation metrics.

Career Highlights

Benz’s listed patent highlights work on systematic evaluation of enterprise RAG systems. The framework’s iterative approach combines performance measurement with automatic parameter adjustment, aiming to help assess and refine these systems.

Collaborations

The provided information names Zhidong Ke and Utsavi Benani as Benz’s coworkers. Their specific contributions to the patent or related work are not specified.

Conclusion

Nicolai Benz’s patent focuses on an iterative framework for measuring and improving enterprise RAG performance, reflecting his work in this area at SAP SE.