LiteParse pairs fast text parsing with a two-stage agent pattern, falling back to multimodal models when tables or charts need visual reasoning ...

Abstract: Visually-situated text parsing (VsTP) has recently seen notable advancements, driven by the growing demand for automated document understanding and the emergence of large language models ...

Swiftlet is a high-performance text-parsing library for Rust, inspired by Python’s Lark. It accepts a context-free grammar (CFG) as input and generates an efficient parser capable of analyzing and ...

In the current landscape of Retrieval-Augmented Generation (RAG), the primary bottleneck for developers is no longer the large language model (LLM) itself, but the data ingestion pipeline. For ...