Skip to content

AI Features

AI4Meta integrates AI throughout the review process to accelerate your work while keeping you in control.

Smart Screening

Our AI screening assistant learns from your decisions and prioritizes studies most likely to be relevant.

  • Active learning — the model retrains after each batch of decisions, surfacing relevant studies first.
  • Confidence scores — each study gets an inclusion probability (0–100%).
  • Suggested decisions — AI pre-labels studies above/below configurable thresholds.
  • Stopping criteria — get notified when the model estimates you've found 95%+ of relevant studies.

You always have the final say — AI suggestions are recommendations, not decisions.

Automatic Data Extraction

Upload PDFs and let AI extract structured data into your codebook fields.

  • Field mapping — AI reads your codebook definitions and extracts matching data.
  • Source highlighting — each extracted value links back to the PDF passage it came from.
  • Confidence indicators — low-confidence extractions are flagged for manual review.
  • Batch processing — extract from multiple PDFs simultaneously.

Improving Accuracy

For best results:

  1. Write clear, specific codebook field descriptions.
  2. Provide examples of expected values in field definitions.
  3. Review and correct the first 5–10 extractions — the AI learns from your corrections.

AI Report Writing

Generate draft sections for your systematic review manuscript.

  • Methods section — describes search strategy, screening process, and analysis methods.
  • Results section — summarizes study characteristics, risk of bias, and meta-analysis findings.
  • Discussion points — suggests interpretation, limitations, and comparison with existing literature.

All AI-generated text is clearly marked as draft. Edit freely before including in your manuscript.

AI Research Assistant

Chat with the AI assistant within any project to ask questions about your data, get help interpreting results, or brainstorm search strategies. The assistant has context about your project's studies and settings.

Data Privacy

AI features process data securely. Your study data is never used to train models or shared with third parties. You can disable AI features entirely in project settings.