Document-aware classification
Identify the document and the job before selecting the execution path.
Agentic Parser for IDP uses the Mozi Model Hub to select an appropriate parsing and reasoning path for each document task.
See Model HubStructure and intent are identified before the task is routed, validated, and returned with reviewable evidence.

Classification, parser choice, extraction, validation, and output evidence remain distinct parts of the job.
Identify the document and the job before selecting the execution path.
Choose a capable, economical model and parser for the specific extraction or reasoning task.
Return structured output with the evidence required for downstream review and audit.