●builderThis benchmark provides a target for developing highly specialized OCR and extraction tools for South Asian legal tech.
●researcherThe dataset reveals significant failures in current multimodal architectures regarding specialized non-standard character sets and positional notation.