AI models for law firms other than LLMs

In this brief dialogue, Speaker A introduces MoneyState, a no‑code platform that lets law firms build AI models and workflows without programming. The speaker argues that generic large language models (LLMs) like ChatGPT often hallucinate legal facts and require extensive rewriting, making them unsuitable for tasks that demand precise, data‑driven results such as summarization, classification, and calculations. MoneyState is positioned as a reliable alternative that extracts accurate information directly from a firm’s own database, avoiding hallucinations. Speaker B clarifies that MoneyState does not replace LLMs entirely; instead, it provides trustworthy structured answers that can later be fed into an LLM (e.g., ChatGPT) for natural‑language drafting. The overall message is that a hybrid workflow—structured AI for factual extraction followed by LLMs for prose—delivers both reliability and efficiency for legal practitioners.










