In the contemporary landscape of technological advancement, artificial intelligence (AI) holds immense potential to address complex issues. This study aims at exploring AI’s applicability in the initial stages of inventive design processes, specifically focusing on the initial situation analysis and resolution phases. It emphasizes leveraging information retrieval techniques, for the efficient design of innovative lattice structures, a noteworthy industrial challenge encompassing mass production at low cost. The approach entails formulating a TRIZ based contradiction query in natural language and implementing an intricate information retrieval pipeline using a meticulously analyzing vast patent databases, as they epitomize a significant knowledge pool in the realm of technological innovations. This research underscores the nuanced challenges associated with optimizing retrieval pipelines to ensure reliable and contextually relevant data extraction. Ultimately culminating in the development of novel lattice structures, our findings accentuate new horizons in industrial production.