Use case
Data extraction
Pulling defined fields out of unstructured text into a schema — from forms, reports, emails, filings or web pages. The general case of invoice extraction, and the step that turns documents into something a database can hold.
- Source
- Editorial ontology entry — no fetched document behind this page
- Evidence
- none on this page — it links to the pages that hold it
- Category
- Documents
- Typical risk
- medium
- Entry
- editorial · reviewed 25 Aug 2026
01What this is
Data extraction takes a target schema and a document and returns validated values. Modern implementations constrain the model to the schema, then validate types, ranges and cross-field consistency before anything is written downstream.
A good deployment defines the schema first, keeps the source span for every extracted value so a reviewer can check it, reports per-field confidence, and refuses to invent a value that is absent — an empty field with a reason beats a plausible guess. Throughput and cost per document are measured alongside accuracy.
Pitfalls: schemas that grow field by field until nothing validates; extraction from documents whose layout carries meaning that plain text loses; and pipelines with no feedback loop, so the same failure recurs monthly. When the source is a third-party website, the terms of that site govern whether extraction is permitted at all.
- Typical data
- forms, reports, email, public filings, scanned documents
- Solution classes it admits
- Enterprise SaaS, Private cloud, Self-hosted
02Deployment options
- ASSESSMENT
On a neutral reading of this use case, Self-hosted is a strong alternative, Private cloud is a strong alternative and Enterprise SaaS is conditional.
- ASSESSMENT
Read as a generic reading of this page, not a recommendation: no organisation, size, jurisdiction, budget or technical capability has been supplied, so wherever an option depends on one of those, it says "unknown". Ask your own question to get a verdict that accounts for them.
Private cloud
STRONG ALTERNATIVEManaged model API or private model deployment in your cloud account
- ASSESSMENT
Two patterns fit inside one account in a region you name: call a managed foundation-model API such as Amazon Bedrock, Microsoft Foundry / Azure OpenAI, or Vertex AI; or deploy an open-weight model on GPU compute you control. Your application, retrieval layer, storage, identity and logs remain in your cloud boundary in both patterns.
- ASSESSMENT
The managed-API pattern can use closed-source frontier models without buying or operating GPUs. It is usually the fastest way to build a custom workflow, but prompts and retrieved context are processed by the managed service, so model availability, retention, abuse monitoring and regional routing must be checked for the exact feature and endpoint.
- ASSESSMENT
The private-model pattern gives more control over weights, serving and network paths, and can use managed endpoints or your own containers. It also makes your team responsible for capacity, patches, model upgrades, evaluation and failover.
- ASSESSMENT
The cloud provider becomes a data processor in either pattern: you need a DPA, a documented region, and an answer on cross-region routing and where support staff can access the environment from.
- RECOMMENDATION
Start with the managed-API pattern when the workflow is custom but model operations are not the source of competitive advantage; move to private model serving only if evaluation, volume, portability or the data boundary justifies the extra operations. Your stated technical capability is "unknown".
Self-hosted
STRONG ALTERNATIVEOpen-weight models on infrastructure you operate
- ASSESSMENT
Documents, queries and embeddings stay on machines you own, using an open-weight model whose licence you review. For a brief that involves confidential documents and personal data, that removes a model-API vendor from the data path rather than governing that transfer by contract.
- ASSESSMENT
It costs you the operational work instead: a GPU server, Docker, Linux, backups and a patching routine. Your stated technical capability is "unknown", which is the attribute this option most depends on.
- ASSESSMENT
No processor agreement, subprocessor list or cross-border transfer assessment is needed for the model itself, because no third party processes the content.
- RECOMMENDATION
Recommended where local processing is preferred (you did not say so) and the content is sensitive (confidential documents and personal data).
Enterprise SaaS
CONSIDER IFFinished closed-source cloud product with enterprise controls
- ASSESSMENT
This is a complete vendor application, not a model API: examples include an enterprise assistant, coding copilot or document product with the workflow, interface, connectors and administration already built. It can use closed-source cloud models while requiring no model hosting or application engineering from your team.
- RECOMMENDATION
Choose it when the product already performs the actual workflow and its controls meet your requirements. Do not choose it only because its underlying model is strong: a finished SaaS product is less flexible than building against a managed API when your process, integrations or review steps are organisation-specific.
- ASSESSMENT
Vendor commitments are treated as unverified until we have fetched the page that makes them. Until then this option carries questions to ask, not assurances: a signed data processing agreement covering the data you will actually put in; a documented data residency commitment naming the region, in the contract rather than a blog post; a written no-training commitment for your content, including uploads and connected sources; stated retention periods and a deletion path you can exercise; an administrative audit log you can export, and SSO with group-based access control.
- RECOMMENDATION
No jurisdiction was named, so this is the check rather than the conclusion: compare the vendor's stated processing locations and subprocessor list against the cross-border transfer rules wherever you operate before uploading anything.
03Deployment stacks
- Self-hostedDocument classification and OCR pipelineDocling (or Unstructured) converts and OCRs, embeddings plus rules classify, a local model reads only the documents that need reading, and Paperless-ngx gives the business a place to search and file. Starts CPU-only; a GPU is added when the volume justifies it.
- Self-hostedContract review and clause extractionDocling converts the contract to structured text with layout preserved; a 32B open-weight model on vLLM fills a JSON schema of clause fields under constrained decoding; results land in PostgreSQL with pgvector for search and precedent lookup; a reviewer approves in the UI.
- Self-hostedInvoice capture and extractionInvoices arrive by email or watched folder, Docling converts them (OCR included), a local vision-capable model fills a JSON schema under constrained decoding, code checks the arithmetic and the supplier against your master data, and n8n posts the clean ones while routing the rest to a person.
04Tools by hosting option
Self-hosted7
- LibraryDoclingDocument conversion toolkit that parses PDF, Office and image files into structured Markdown or JSON, preserving reading order, tables and figures for downstream retrieval.
- FrameworkLlamaIndexData framework for retrieval applications: loaders for many document types, indexing and query pipelines, and evaluation helpers for retrieval quality.
- HybridMilvusDistributed vector database designed for large collections, with several index types, GPU indexing options and a separated storage and compute architecture.
- LibraryPaddleOCROCR toolkit with detection, recognition and layout models, including Chinese and other East Asian scripts, plus table and formula recognition. Runs offline on CPU or GPU.
- Self-hostedRAGFlowRetrieval engine built around deep document parsing: layout-aware chunking of PDFs, tables and scans, citation-backed answers, and a visual pipeline for building knowledge bases.
- LibraryTesseract OCRLong-established open-source OCR engine with trained data for many languages and scripts, usable offline and embedded in most self-hosted document pipelines.
- HybridUnstructuredLibrary and hosted API that partition documents of many formats into typed elements for indexing, with connectors to common storage systems and vector databases.
Private cloud11
- SaaSAmazon TextractAWS service extracting text, forms and tables from scanned documents, with specialised APIs for invoices, receipts and identity documents, priced per page.
- SaaSAzure AI Document IntelligenceAzure service for OCR, layout analysis, prebuilt document models and custom extraction, deployable in a chosen Azure region and available as a container for local processing.
- LibraryDoclingDocument conversion toolkit that parses PDF, Office and image files into structured Markdown or JSON, preserving reading order, tables and figures for downstream retrieval.
- SaaSGoogle Document AIGoogle Cloud service for OCR, form parsing and specialised document processors, with regional processing options and integration into the wider Google Cloud data stack.
- PlatformGoogle Vertex AIGoogle Cloud platform for Gemini and third-party models with regional endpoints, VPC Service Controls, grounding against your own data and enterprise IAM.
- FrameworkLlamaIndexData framework for retrieval applications: loaders for many document types, indexing and query pipelines, and evaluation helpers for retrieval quality.
- HybridMilvusDistributed vector database designed for large collections, with several index types, GPU indexing options and a separated storage and compute architecture.
- SaaSNanonetsHosted document AI service with trainable extraction models, an approval interface and workflow integrations for accounts payable and back-office document types.
- Self-hostedRAGFlowRetrieval engine built around deep document parsing: layout-aware chunking of PDFs, tables and scans, citation-backed answers, and a visual pipeline for building knowledge bases.
- SaaSRossumCloud platform for transactional document processing — invoices, orders, delivery notes — with extraction, a validation interface for exceptions and ERP integrations.
- HybridUnstructuredLibrary and hosted API that partition documents of many formats into typed elements for indexing, with connectors to common storage systems and vector databases.
Vendor cloud9
- SaaSAmazon TextractAWS service extracting text, forms and tables from scanned documents, with specialised APIs for invoices, receipts and identity documents, priced per page.
- SaaSAzure AI Document IntelligenceAzure service for OCR, layout analysis, prebuilt document models and custom extraction, deployable in a chosen Azure region and available as a container for local processing.
- SaaSGoogle Document AIGoogle Cloud service for OCR, form parsing and specialised document processors, with regional processing options and integration into the wider Google Cloud data stack.
- PlatformGoogle Vertex AIGoogle Cloud platform for Gemini and third-party models with regional endpoints, VPC Service Controls, grounding against your own data and enterprise IAM.
- HybridMilvusDistributed vector database designed for large collections, with several index types, GPU indexing options and a separated storage and compute architecture.
- SaaSMindeeDocument parsing API with prebuilt models for invoices, receipts and identity documents, plus custom model training. European vendor with a developer-first API.
- SaaSNanonetsHosted document AI service with trainable extraction models, an approval interface and workflow integrations for accounts payable and back-office document types.
- SaaSRossumCloud platform for transactional document processing — invoices, orders, delivery notes — with extraction, a validation interface for exceptions and ERP integrations.
- HybridUnstructuredLibrary and hosted API that partition documents of many formats into typed elements for indexing, with connectors to common storage systems and vector databases.
On-premise (enterprise plan)3
- SaaSAzure AI Document IntelligenceAzure service for OCR, layout analysis, prebuilt document models and custom extraction, deployable in a chosen Azure region and available as a container for local processing.
- LibraryPaddleOCROCR toolkit with detection, recognition and layout models, including Chinese and other East Asian scripts, plus table and formula recognition. Runs offline on CPU or GPU.
- LibraryTesseract OCRLong-established open-source OCR engine with trained data for many languages and scripts, usable offline and embedded in most self-hosted document pipelines.
05Compliance hot spots
This use case usually raises personal data, terms-of-service restrictions, retention, auditability, data residency, automated decision-making, confidentiality, logging, sector rules, security.
- Terms-of-service restrictions
- No published jurisdiction page names this topic yet
- Retention
- Hong Kong
- Auditability
- No published jurisdiction page names this topic yet
- Data residency
- No published jurisdiction page names this topic yet
- Automated decision-making
- European UnionUnited KingdomSouth Korea
- Logging
- European Union
- Sector rules
- Hong Kong
- Security
- China (mainland)
06Example questions
Each of these opens the question box with the text already in it. The answer is researched for your organisation, not for this page.
07Related use cases
- DocumentsSpreadsheet analysisAsking questions of tabular data in natural language — totals, trends, outliers, reconciliations — and getting an answer with the calculation behind it. Works best when the model writes code that runs, rather than reading numbers itself.
- DocumentsRFP draftingAnswering tenders and security questionnaires from a library of previous answers. The work is retrieval plus reuse: most questions have been answered before, in a form somebody already approved.
- DocumentsReport generationProducing a recurring document — a management pack, client update, board summary, compliance return — from data and prior text. The template and the data are fixed; the narrative around them is what takes the time.
- DocumentsOCRTurning images of text — scans, photographs, faxes, historic files — into machine-readable text with layout. The step everything else depends on: no extraction, search or classification pipeline is better than the text layer underneath it.
- DocumentsInvoice extractionReading supplier invoices — header fields, line items, tax, totals — into structured records. The extraction step only. What happens to the record afterwards is invoice processing automation.
- DocumentsDocument classificationSorting incoming documents into types and routing them — which team, which folder, which workflow, which retention rule. Usually the first step of a larger pipeline and the cheapest place to remove manual handling.
Improve this page
Sign in to contribute
From the field
0 deployments · 0 questions
Nobody has reported deploying this here yet, and no question has been opened against this page. Both appear once a reviewer accepts them.