Documents Overview
Muin Documents is one AI-powered library for every business document: upload anything, get structured data, find it instantly, and catch problems early.
Muin Documents is the platform’s foundation: a single, AI-powered library for every file your organization touches, invoices, contracts, receipts, certificates, statements, and everything in between.
Why Use Documents?
- Stop retyping paperwork. Upload a document and Muin’s AI extracts the structured data, vendor, amounts, dates, line items, in seconds, with confidence scores you can review.
- Find anything instantly. Keyword and semantic search work across the content of your documents, not just file names. Ask for “the contract with the 90-day termination clause” and get it.
- Catch problems before they cost money. The Second Pair of Eyes engine flags duplicates, unusual amounts, contract mismatches, and policy violations automatically.
- Feed every other module. Extracted document data flows into Finance (bills, expenses), Vendors (certificates, compliance), HR (employee records), and your workflows: entered once, used everywhere.
Capabilities
| Capability | What It Does | Guide |
|---|---|---|
| Uploading | Drag-and-drop, bulk upload, email-in, and mobile capture | Uploading Documents |
| AI Processing | Type detection, field extraction, confidence scoring | Processing & AI Extraction |
| Search | Keyword, semantic, and advanced filtered search | Searching & Filtering |
| Organization | Folders, tags, categories, smart collections | Organizing Documents |
| Templates | Generate documents from templates with auto-filled data | Document Templates |
| Anomaly Detection | Duplicates, amount anomalies, policy violations | Second Pair of Eyes |
| Troubleshooting | Fixes for upload, processing, and display issues | Troubleshooting |
Start Here
- Upload your first document: see extraction work on a real invoice in ~30 seconds
- Understand AI processing: what the confidence scores mean and when to review
- Organize as you grow: folders and tags that keep a large library navigable
Related Features
- Document Assembly: compose new documents from reusable building blocks
- Workflows: route documents for approval automatically
- E-Signatures: built-in acknowledgments today; legally binding e-signatures are not yet available
Document types sit behind the library
Every document has a type, and the type is what decides which fields Muin extracts, which keywords find it in chat, and the one workflow that governs its approval and lifecycle. Each type also has its own hub page gathering its documents, templates, Composer projects and approval policy. Owners and admins manage types at Administration -> Document Types. See Document Templates.
Checking extraction quality across the library
Administration -> Processing reports the pipeline for the whole organization: how many documents are pending, processing, completed and failed, plus Extraction Confidence, Fields Extracted, Search Chunks and Review Rate, broken down by document type, with the documents that need attention listed. See Platform Console.
Administration -> Search Analytics reports whether people are finding things: total searches, average results per search and the zero-result rate, with the top queries. A high zero-result rate usually means titles are not descriptive or scans have no text layer.
When it goes wrong
| What you see | What to do |
|---|---|
| Extraction returned nothing useful | Check the document type is right, and that the file has a text layer rather than being a flat scan. See Troubleshooting. |
| A document has no governing workflow | Its type has none attached. Fix it at Administration -> Document Types. |
| Nobody can find a document by searching | Give it a descriptive title and tags; check Search Analytics for the exact failed query. |
| An upload never finished | Check Administration -> Async Failures for the background job, and requeue it there. |
Need Help?
Ask Muin Assistant (“How does document extraction work?”) or reach support via Help -> Contact Support.