Nested Knowledge is a systematic review and evidence synthesis platform built to compress months of review work into weeks — with AI features that are traceable, validated, and aligned to Cochrane’s RAISE recommendations for responsible AI in evidence synthesis.
In an independent evaluation across live projects at Certara, automation in Nested Knowledge reduced workload by an average of 85% in key review steps, while maintaining or surpassing human-level accuracy. — Grys et al., AI in Evidence Synthesis: Have We Reached the Promised Land or Are We Still Wandering the Desert? ISPOR Europe 2025, HTA29 (poster PDF).
See below for an overview of Nested Knowledge’s capabilities:
Why teams choose Nested Knowledge #
Cut review time without cutting corners. AutoLit® compresses search, screening, tagging, extraction, and critical appraisal into a single connected workflow. Every step that can be automated has an AI-assisted option; every AI output is traceable to its source and reviewable by a human before it becomes part of your review.
- 85% average workload reduction in independent third-party evaluation: Certara’s ISPOR Europe 2025 evaluation across live projects found automation in Nested Knowledge cut workload by an average of 85% in key review steps. Criteria-Based Screening saved ~80% of time at ~95% accuracy; Adaptive Smart Tags extracted structured data at 90–100% accuracy with 80–90% time savings; automated critical appraisal saved ~100 hours in a single 280-study project. See Validation Studies of AI Tools in Nested Knowledge for the full breakdown, and Artificial Intelligence in Nested Knowledge for the feature inventory behind it.
- RAISE-aligned by design: Our AI features map against Cochrane’s RAISE recommendations for responsible AI in evidence synthesis: traceability on every AI output, in-review validation, hybrid rule-based and LLM approaches where appropriate, and human oversight at every decision point. Learn more about how Nested Knowledge is aligned with RAISE here.
- Independently validated: Every Smart tool ships with published or in-house validation data, and the Certara evaluation above provides third-party benchmarking against manual methods. Per-model accuracy, sensitivity, and specificity metrics are available on each model card.
- Human oversight: Single or dual screening, dual critical appraisal, and full audit trails are built in. AI accelerates the workflow; reviewers make every final call.
The AutoLit workflow #
| Stage | What it does | AI assistance |
| Search | Build a structured search with Search Exploration, run automatic and updatable searches on PubMed, or import records from other indices. | Smart Search |
| Screen | Configure exclusion reasons and include/exclude studies via single or dual screening. OR Create a set of yes/no criteria via Criteria-based Screening. | Robot Screener, Smart Screener |
| Tag | Build a tag hierarchy of concepts, or import an organisational template and extract data. Produces a Qualitative Synthesis diagram automatically. | Adaptive Smart Tags, Core Smart Tags |
| Meta-Analytical Extraction (optional) | Capture quantitative data that auto-populates Quantitative Synthesis with a network meta-analysis. | Smart MA Extraction |
| Critical Appraisal (optional) | Run Risk of Bias and quality assessments, single or dual. | Smart Critical Appraisal |
| Outputs | Compose your Abstract, full Manuscript, and Dashboard of visuals and Insights. | Smart Insights |
Throughout, Study Inspector is your single view of every study in the nest — filter, edit, and export from one place.
Get started #
- Getting Started guide — step-by-step walk-through of setting up your first nest.
- Demo Nests — try a real, editable project: CER (aspiration catheter), HTA (TKIs for NSCLC), or ACE inhibitors & ARBs for heart failure.
- Sign up — or contact us for a guided demo.
Responsible AI & validation #
For teams whose AI use needs to stand up to internal review, regulator scrutiny, or journal expectations:
- Artificial Intelligence in Nested Knowledge — full catalogue of AI features by review stage.
- Validation Studies of AI Tools in Nested Knowledge — published validation data for every Smart tool.
- Crafting the Perfect Prompts for Nested Knowledge’s AI Tools — guidance on prompt design.
- Guidance on using AI in Nested Knowledge — when to use which tool, and how to document AI use in your protocol.
- Disclosure of AI Systems and AI Compliance with Legislation — formal policy documents.
- National Academies webinar with Kevin Kallmes, CEO — AI in Evidence Synthesis. Watch the webinar.
Model cards #
Technical specifications for every AI model in the platform: Smart Study Type · Smart Study Size · PICO Hierarchy Generator · Smart Study Location · BioELECTRA-PICO · Smart Search · Research Question Refinement · Screening Model & Robot Screener · Adaptive Smart Tags · Smart Critical Appraisal.
Methodology resources #
- Systematic Review Guide and Performing a Meta-analysis — methodology fundamentals.
- Best practices for search, screening, tagging, meta-analytical extraction, and critical appraisal.
- Rapid Reviews in Nested Knowledge and Guidance on Review Types — choose the right review design.
- Writing for publication: finding a journal, manuscript writing, formatting, submission tips.
- Free course: How to Review the Medical Literature.
Support #
- Ask AI ChatBot — first stop for product questions.
- Support page or email support@nested-knowledge.com for technical issues.
- Suggest an improvement
