AI glossary builder
Generate a glossary from any document — every term cited.
Dense documents bury their jargon. Sidenote reads the whole thing and builds a glossary of the acronyms, terms and proper nouns — each defined in plain English and citing the exact passage it came from. Click any entry to scroll straight to where the term first appears.
The problem
The jargon is the part that slows you down.
Every field has its own shorthand. A research paper drops a dozen acronyms before page two; a runbook assumes you know the service names; an onboarding doc is half abbreviations. You can read the words and still not follow the argument, because the vocabulary was never explained.
A glossary fixes that — but building one by hand means re-reading the whole document and hunting for first mentions. Sidenote does that pass for you, and ties each definition back to the line it came from so you can check it in a click.
How it works
How Sidenote builds the glossary.
What it extractsAcronymsThe abbreviation plus its expansion — “GLP-1” becomes “Glucagon-like peptide-1”, pulled from the document's first mention. Domain jargonThe field-specific vocabulary that carries the argument — defined in one sentence, grounded in how the document uses it. Proper nounsThe people, products, places and named entities the document references — so you know who and what it's talking about.
Acronyms, jargon and proper nouns.
Sidenote sorts each term by kind, so a glossary reads the way you'd expect — the abbreviation tells you its expansion, the entity tells you who or what it is. Common words and bare numbers are left out unless the document gives them a specific meaning.
Across many documentsMerged and deduplicatedA term that appears in three papers shows once — not three near-identical entries cluttering the list. Each entry names its documentEvery definition tells you which document it came from, so you can trace it back across the whole Collection. Reuses prior workDocuments you've already glossarised are reused from cache; only the new members trigger a fresh build.
One glossary across a whole Collection.
A single document's glossary is useful; a shared vocabulary across a project is better. Bundle your papers, wiki pages or specs into a Collection and Sidenote merges a single, deduplicated glossary across all of them — each entry still naming the document it came from, so the citationtrail never breaks. It reuses any per-document glossary you've already built, so only the new documents do fresh work.
Everywhere you read
Glossarise every kind of document.
The same builder works wherever Sidenote reads — browser PDFs and articles, your wikis, your docs, and files you upload. Pick a source:
FAQ
Common questions about the glossary builder.
How do I generate a glossary from a document?
Open Sidenote on any document you're reading — a PDF, a wiki page, a paper, a Google Doc — and switch to the Glossary tab. Click Build and Sidenote reads the full text, extracts the acronyms, jargon and proper nouns, and writes a one-sentence definition for each. There's nothing to upload and nothing to paste: it works on the document where it already lives.
Does each definition cite where the term came from?
Yes. Every glossary entry carries the passage it was found in. Click the entry and Sidenote scrolls the live document to the first place the term appears and highlights it — the same scroll-to-source behaviour as a chat citation. A definition you can't trace back to the text isn't much use, so each one is anchored to its source.
Where do the definitions come from — the document or general knowledge?
The document first. Sidenote uses any first-mention expansion the text gives you (“Upper motor neurons (UMNs)” becomes “UMN” defined) and prefers phrasing grounded in the passage. For widely-known terms the document uses without re-defining — RNA, OAuth, NASA — it falls back to a short general-knowledge definition so the term is still explained. It only omits a term when it's both opaque in the document and not something it can define confidently.
Can I build a glossary across several documents at once?
Yes — that's a Collection glossary. Bundle a set of related documents into a Collection and Sidenote merges a single, deduplicated glossary across all of them, with each entry still naming the document it came from. It reuses any per-document glossary already built, so it only does fresh work on the documents that need it.
What kinds of terms does it pick up?
Three kinds: acronyms (with their expansion), domain jargon, and proper nouns — people, products, places and named entities. Common words like “data” or “system” are skipped unless the document gives them a specific meaning, and stop-words, dates and bare numbers are never defined.
Which documents can I glossarise?
Everything Sidenote reads: browser PDFs and web articles, Confluence, Notion, Google Docs, SharePoint and OneDrive, and files you upload (PDF, DOCX, PPTX, TXT, MD). Scanned PDFs are OCR'd first, so even an image-only document can be turned into a glossary.
Try it on your next dense doc
Decode the jargon. In one click.
Add Sidenote to Chrome, open a document, and hit Build on the Glossary tab. Watch it pull out every term — each one cited to the line it came from.
7-day Pro trial · No card required · Free tier forever