LandingAIOpen Playground

Five tools, one workflow

Classify, Parse, Section, Split, and Extract — when each runs and what it produces.

This module

Understand the ADE toolkit. know which tool to reach for and when.

Classify filters and routes pages before parsing. Parse is the foundation. Section, Split, and Extract operate on the parsed output and serve different downstreams.

Raw documentsPDFs · images · scansPRE-PARSE1. ClassifyLabel each page per class —filter or route to pipelines.FOUNDATION2. ParseDocument → JSON + Markdown.Layout, hierarchy, and visualgrounding all preserved.FOR RETRIEVAL3a. SectionBuild a TOC with the hierarchylevel of each section.Attach as chunk metadata.FOR EXTRACTION3b. SplitSeparate by document typeand identifiers withinthe same type.DOWNSTREAMRetrieval / RAGHierarchy-aware chunks feedyour vector DB or search index.FOR EXTRACTION4. ExtractPer-type schema extraction —a different schema for eachdocument type.DOWNSTREAMCSV or your databaseExport rows for analysis,or load straight into your DB.

What you can build

Concrete use cases for each tool.

Parse
RAG over a PDF library

Turn thousands of reports, manuals, and filings into clean, chunk-ready Markdown for your vector DB.

Extract
Invoice processing

Pull vendor, line items, and totals from invoices as structured JSON ready for your ERP or accounting system.

Split
Mortgage packet triage

Break a 50-page loan packet into separate pay stubs, bank statements, and tax forms — each routed to the right workflow.

Documentation · Playground · Pricing