docRefinery vs Extend

Two products that both get better with use

This is the closest comparison on our list, because the feature lists overlap: extraction, classification, splitting, confidence scores, human review, versioning. The difference is what improves. Extend's Studio is built for iterating on a schema, running evals and catching regressions, and its Composer agent refines schemas automatically. docRefinery keeps the sender's layout itself. Theirs tunes the question against a test set. Ours remembers the page.

Neither approach is wrong. A varied document set with a dedicated AI engineer suits the first. A finance team receiving the same few hundred layouts every month suits the second.

Where Extend is the better choice

• Proving quality over time. If a stated requirement is to show how a pipeline's accuracy moves release by release, Extend puts evaluation at the centre of the product. Our versions and replay help, but we have no standalone evaluation harness and no published accuracy figure.
• Regulated buyers. Extend says it is certified for SOC 2, HIPAA and GDPR and undergoes regular third-party penetration testing. A BAA is an add-on even on Pay As You Go. docRefinery's SOC 2 Type II audit is in progress with no report yet, and we make no data-residency commitments, so a buyer who needs either today should pick Extend. Ask us for the latest status.
• EU hosting and deployment choice. Extend offers EU data residency on all plans, plus bring-your-own-cloud and hybrid set-ups where documents stay in your cloud.
• Self-serve pricing. Credits are published: Extract runs at 3 credits a page in Performance mode and 0.7 in Light. You can estimate a bill without a call.
• Fraud checks. Verify flags AI-generated documents and signs of tampering. We do not offer this; forgery detection sits on our roadmap as research, which means it may not ship.
• File breadth. Extend lists 35+ file types. Presentations and wider office formats are in development at docRefinery and not yet available.

What docRefinery does that Extend does not, or does differently

docRefinery fits when the input is repetitive and the people running it are operations staff, not ML engineers. Against Extend, these are the differences that show up in daily work.

• Layout memory, not schema tuning. Each sender's layout is learned once and remembered, and re-trained when a form changes, so improvement comes from the documents you already process rather than from an eval set someone maintains.
• You define the output once. Each field has a plain-language instruction, a "Return as" name you choose as its output key, a type and a required flag, held in account configuration rather than passed in each call.
• One output format for everything. Every field returns text, alias, field_confidence and location, and every table returns rows, so one parser in your code handles every document type.
• Governed fields. Owner, admin and member roles decide what may be extracted, and a caller cannot override it. Extend lets developers choose inline or saved configuration.
• Null, never a guess. A field missing from the page comes back null, and anything under its per-field threshold waits in a review queue for a person.
• Corrections that become fixes. A reviewer's correction becomes a suggested fix, tested on that layout's own documents before anyone applies it.
• A workflow from one document. Drop one sample and every printed field is found, ready to keep, relabel or switch off. Business rules, such as line items adding up to the total, run on every document, and each change is a version you can replay against past runs. See it on invoice data extraction and receipts.
• Long files, segmented. Long, multi-page files are handled at scale and separated into the documents inside them, each classified into its type.
• Air-gapped deployment. Beyond our cloud and your private VPC, docRefinery runs fully air-gapped on your own servers on Enterprise. Extend lists its cloud, bring-your-own-cloud and hybrid.
• Plug and play. API keys with expiry, revocation and an activity log, export to JSON, CSV or XLSX that always reflects corrections, a Chrome extension in early access that fills web forms from a document, and custom document types trained and hosted at no charge.