Literature review

A literature review asks you to do something your brain isn't built for: hold dozens or hundreds of papers in your head at once, notice where they agree, where they contradict, and where the gaps are. You can read them one at a time. You can take notes on each. But the synthesis, the thing that makes a literature review more than a list of summaries, depends on seeing across the whole body of work at once. That's where most people's systems fall apart.

This page is for researchers, postgraduates, and students doing a literature review and looking for a way to collect, annotate, and search across a large body of papers, with an AI that helps surface the connections between them.

The problem

The volume defeats you. A serious literature review might cover fifty, a hundred, even several hundred papers. You read each one carefully, but by paper forty you've lost the detail of paper twelve. The earlier sources blur, and the connections that should emerge between them stay invisible because you can't hold enough in working memory at once.

Your notes and annotations are scattered. Some highlights live inside a PDF reader, some notes are in a document, some are scribbled in margins you'll never search again. When you need to pull together everything you've read about a specific concept, you're looking in five places and hoping you don't miss something.

Synthesis depends on memory, and memory fails. The whole point of a literature review is to find patterns: agreement, contradiction, evolution of an idea, gaps nobody has addressed. Those patterns exist across papers, not within them. And they only become visible when you can see your whole reading at once, which no folder of PDFs lets you do.

What Fabric changes

Every paper and every note is searchable by meaning. Search for a concept in plain language and Fabric finds every paper, annotation, and note that touches it, across your entire collection. You're not relying on your memory of what you read or your ability to guess the right keyword. The full body of work is queryable.

Your annotations stay alive and findable. Highlights, comments, and marginal notes made on any paper become part of the searchable whole. When you're writing up and need to pull together what you thought about a concept across twenty papers, those annotations come back, not locked inside individual PDFs but surfaced alongside everything else.

The connections become visible. Because Fabric understands what your papers are about, it can surface relationships between sources that you wouldn't see by reading them sequentially. Two papers from different subfields making the same claim, an argument evolving over a decade, a gap sitting between established positions.

How it works

Search across your entire reading. Fabric's AI search works on meaning, not keywords, and reads inside every PDF, document, and note. Ask "critiques of dual-process theory in decision-making" and get every relevant paper and annotation, however they were titled.

An AI that reads with you. The AI assistant works from your saved papers and notes, so you can ask it to summarise what your sources say about a theme, identify disagreements between authors, or find papers in your collection that address a specific question. It synthesises from your reading, not the open internet.

Annotate directly on papers. Annotate PDFs and readings with highlights and comments, and those annotations become searchable. Your thinking while reading is as findable as the papers themselves.

Write alongside your sources. Use notes and docs to draft sections of your review with your annotated papers right there. Link to and embed sources as you write, so the review stays grounded in the material.

Capture papers from anywhere. Clip from the web, save from databases, forward a paper to your email-to-note address, or drop a PDF in directly. Every paper ends up in the same searchable collection regardless of where you found it.

A literature review workflow in Fabric

Build a space for the review. Create a space for the project so the whole body of reading is grouped together, browsable when you want focus, searchable when you want breadth.

Save and annotate as you read. Every paper goes into Fabric with your highlights and notes. Don't organise yet. Just capture your thinking while it's fresh.

When you've read enough, ask. Search for a theme across all your papers at once. Ask the assistant what your sources collectively say about a concept. Ask where they disagree. Ask what none of them address. These are the synthesis questions that take hours manually and seconds with a searchable body of work.

Draft from the synthesis. Write up sections of the review in Fabric, pulling in sources and annotations as you go. When you need to check a claim or find where an idea first appeared, search rather than scroll.

Fill the gaps. The process of searching and synthesising reveals what's missing. The question nobody answered, the method nobody tried, the contradiction nobody reconciled. That's where your contribution lives.

What compounds over time

A literature review that lives in Fabric doesn't end when the paper is submitted. The collection of annotated sources stays, searchable and connected, so when you return to the topic for a later project, a conference paper, or a new chapter, the foundation is already there. Every paper you add deepens what the AI can do. A collection of two hundred annotated papers is qualitatively different from twenty, not just in volume but in the connections it can surface.

Researchers who maintain their reading in Fabric across projects find that reviews get faster the second and third time, because the work of reading and annotating isn't thrown away between projects.

For a structured approach, see the guides to literature reviews and research workflow.

Related use cases

A literature review often lives inside a larger dissertation or thesis, and the skills transfer directly to broader research projects. If you're building a permanent personal knowledge system beyond any single review, see second brain. For the reading-and-saving layer specifically, reading and learning covers how Fabric works as a personal library. Fabric is built for researchers and students at every stage.

Get started

Bring your papers, annotations, and notes into one place and run a literature review from a system that sees across all of them. Try Fabric free.

Comparing tools? See why researchers choose Fabric as the best way to organise research and the best app for PhD students.

FAQs

How many papers can Fabric handle for a literature review? There's no practical limit. Whether you're working with fifty papers or five hundred, every one is searchable by meaning and included in what the AI can draw on. The system is built for the scale a real literature review demands.

Can I search for a concept across all my papers at once? Yes. Search in plain language and Fabric finds every paper, annotation, and note that touches the concept, regardless of how individual papers are titled or which keywords they use. It searches by meaning, not by filename or exact phrase.

Can the AI summarise what my sources say about a specific topic? Yes. Ask the assistant about any theme and it synthesises from your saved papers and notes. It can summarise agreement, highlight disagreements between authors, or identify what none of your sources address, all from your collection, not the open web.

Do my annotations on PDFs become searchable? Yes. Highlights, comments, and marginal notes made on any paper in Fabric become part of the searchable whole. When you search a concept, your annotations come back alongside the source text.

Can I save papers from Google Scholar, PubMed, and university databases? Yes. Clip papers from the web, save downloaded PDFs, or forward them to your email-to-note address. However a paper reaches you, it arrives in the same searchable collection.

Does it replace Zotero or Mendeley? Fabric isn't a citation-formatting tool. It complements a reference manager. You keep Zotero or Mendeley for generating bibliographies and use Fabric as the place where you read, annotate, search, and synthesise across your papers.

Can Fabric help me find gaps in the literature? The AI can show you what your collected sources address and where they don't converge. Asking "what don't any of my papers address about X" or "where do my sources disagree" is a direct way to find gaps. The more papers you've collected and annotated, the more useful this becomes.

Can I organise papers by theme and still search across all of them? Yes. You can group papers into sub-spaces by theme, method, or chronology, and still search across the whole collection. Focused views and broad search coexist.

Can I share my literature review collection with a supervisor or collaborator? Yes. You can share a space or document with specific people, or publish with password protection and analytics so you control who sees what and can track engagement.

How is this different from using a folder of PDFs? A folder stores files by name. Fabric reads the content of every paper, understands what it's about, indexes your annotations, and lets you search and ask questions across the whole body of reading by meaning. The difference is the gap between storing papers and being able to synthesise from them.

Can I use Fabric for a systematic review with specific inclusion criteria? Fabric supports the reading, annotating, and synthesis stages of a systematic review. You can tag papers by inclusion status, annotate them against your criteria, and search across included papers for synthesis. The protocol and reporting framework are yours; the searchable, annotated collection is Fabric's contribution.

Will my literature review collection be useful for future projects? Yes. The collection stays, searchable and growing, so returning to a topic for a later paper, a new chapter, or a conference submission means the foundation is already there. Papers and annotations from one review are immediately available in the next.