There is an emerging technology called large language models (LLMs) that can help with intelligent document processing by extracting information from large volumes of unstructured data. This technology is being used in deepset Cloud, a platform that allows cross-functional AI teams to build shippable products. LLMs are particularly useful for tasks such as due diligence processes, compliance reviews, and audit activities, where manual extraction of structured information from unstructured data is required. The platform provides features such as query templates, iterative refinement, and testing to help users optimize their LLM setup configuration, prompt writing, and hyperparameter tuning. Once set up, the platform can generate memos containing relevant data extracted from underlying documents, which can be easily fact-checked by human analysts and shared with other stakeholders. By automating repetitive tasks, LLMs enable organizations to streamline document set analysis, freeing up subject matter experts to apply critical thinking skills to model results.