MetaBeeAI#
MetaBeeAI is an open-source, modular pipeline for extracting structured information from scientific papers for systematic review and meta-analysis in biology.
Use these docs to install MetaBeeAI, configure reproducible processing runs, understand the major modules, and navigate the full workflow from raw papers to structured outputs.
Start with Installation and Quick Start. If you already have the package running, jump straight to Configuration, Workflow, or Reference.
Explore The Docs#
Set up the package, configure required keys, and verify that the CLI is available.
Follow the shortest path from installation to processing a set of PDFs.
Read the end-to-end guides for setup, configuration, workflow, benchmarking, and troubleshooting.
Find API docs, submodule overviews, and lower-level reference material in one place.
Find contribution, configuration-development, and documentation-maintenance guides for working on MetaBeeAI itself.
Pipeline Guides#
Prepare local directories, expected file layout, and the project structure used by the pipeline.
Understand configuration files, environment variables, defaults, and precedence.
Follow the end-to-end processing flow from raw PDFs to structured outputs.
Compare model behaviour and evaluate extraction quality across runs.
Work with processed outputs and downstream analysis steps after extraction is complete.
Diagnose installation issues, pipeline failures, and common runtime problems.
Reference And Internals#
Read higher-level descriptions of the major pipeline components and what each one is responsible for.
Understand how PDFs are split, merged, deduplicated, and prepared for downstream extraction.
Review the prompts, extraction flow, and higher-level orchestration behind the LLM stage.
Find the query and analysis layer used for working with processed outputs.
Development#
Set up a development environment, run checks, and navigate the project’s developer documentation.
Extend the configuration layer safely and consistently.
Build, debug, and maintain the Sphinx documentation stack.