Blog

Technical

Document Processing APIs for SaaS: A Practical Guide (2026)

Your SaaS customers want document upload. A document processing API lets you ship extraction without building it from scratch. Here's what to look for.

Uri MerhavUpdated Mar 26, 2026 · 8 min

Company

One Row Per Form: Your Browser Can Read a Filled PDF, Acrobat Charges $240 a Year for a Worse Version

Fifty filled forms come back and the real job is lining them up: one row per form, one column per field. Acrobat can merge form data into a spreadsheet, but it demands identical forms, aborts on one bad file, and writes empty CSVs near a hundred files. We built a free tool that reads digital forms in your browser - no upload - and falls back to AI only for scans. This post shows a real filled form and the grid it became.

Uri MerhavUpdated Jul 6, 2026 · 6 min

Company

The Auto-Renewal Tripwire: Term-Switcheroos, Notice Ceilings, and Certified-Mail Clauses

A contract does not renew by accident - someone wrote the clause to renew. Here is how auto-renewal clauses are built to slip past you, and a free tool that reads the notice deadline out of every contract and drops it on your calendar.

Uri MerhavUpdated Jul 6, 2026 · 6 min

Company

10,000 Free Downloads a Day: Why We're Doubling Down on Free Tools

Every day, people pull more than 10,000 free downloads out of DocuPipe - spreadsheets, redacted PDFs, split files, searchable scans - most of them without ever creating an account. A founder's note on why that number makes us happy, and why we just shipped four more free tools.

Uri MerhavUpdated Jul 5, 2026 · 6 min

Company

Paste a Screenshot, Get a Workbook: Free Image to Excel Without the Retyping

We built a free image to Excel tool around the fastest input there is: Ctrl+V. Paste a screenshot or a phone photo of a table and get an editable workbook - every table its own sheet, numbers typed as numbers. This post shows the actual photo and the actual rows that came out.

Uri MerhavUpdated Jul 5, 2026 · 5 min

Company

Five Shops, Five Email Layouts, One Tracker: Free Email to Excel With Nothing to Authorize

Email parsers want mailbox access and template setup before they show any value. We built a free email to Excel tool that reads only what you paste: each email becomes a row in one tracker spreadsheet. This post shows the actual raw emails - including a forwarded one with quoted history - and the tracker they became.

Uri MerhavUpdated Jul 5, 2026 · 6 min

Company

Visually a Table, Structurally Div Soup: Free HTML to Excel for the Modern Web

HTML-to-Excel converters map <table> tags one to one - and most of the modern web stopped using table tags years ago. We built a free tool that reads what the markup looks like, not what it's tagged as. This post shows a supplier directory with zero table tags in the source and the 15 clean rows that came out.

Uri MerhavUpdated Jul 5, 2026 · 6 min

Company

The Tables Word Never Told You About: A Free Word to Excel Converter That Reads Like a Person

Half the tables in real Word documents are not table objects at all - they're tab-aligned text. Mechanical converters silently drop them. We built a free Word to Excel tool that reads the document like a person does, and this post shows the actual document where Word says there's one table and the workbook proves there are two.

Uri MerhavUpdated Jul 5, 2026 · 6 min

Company

A Black Rectangle Is Not a Redaction: A Free PDF Redactor Built Around Review

We built a free PDF redaction tool around the two things that actually matter: a review screen that shows every proposed box on the real page, and a download that is burned in - pixels masked, text layer removed - not a cosmetic overlay.

Uri MerhavUpdated Jul 2, 2026 · 6 min

Company

Loss Runs to Excel, Rent Rolls to Excel: Two Free Tools for the Worst Retyping Jobs

We shipped two free tools: drop loss run PDFs from every carrier and get one consolidated claims workbook, or drop any rent roll and get a standardized Excel with a totals tie-out check. In the browser, no signup.

Uri MerhavUpdated Jul 2, 2026 · 6 min

Company

Search the Pile, Split the Pile: Two Free Tools for Scanned Paper

Two free browser tools for scanned paper as it actually arrives: search a whole pile of image-only scans at once like a search engine, or split one stapled-pile PDF at AI-detected boundaries with every piece named by its content. No signup.

Uri MerhavUpdated Jul 2, 2026 · 6 min

Company

Vendor Paperwork Is a Matching Problem: Two Free Tools for Quotes and COIs

Two free browser tools for vendor paperwork: align 2-6 vendor quotes into one comparison matrix even when line items are worded differently, and turn a folder of certificates of insurance into one tracker with expired and expiring-soon highlighting. No signup.

Uri MerhavUpdated Jul 2, 2026 · 6 min

Company

DocuPipe for n8n: Document Extraction Inside Your Workflows

We just shipped n8n-nodes-docupipe, a community node that drops DocuPipe's document extraction straight into your n8n workflows. Invoices, contracts, medical records, handwriting, 150+ languages, any layout. One node away from the rest of your stack.

Ben SaadUpdated Apr 22, 2026 · 8 min

Technical

The Deployment Reality: Cloud vs. VPC vs. On-Premise Document AI

Multi-tenant document AI fails government and healthcare audits. Learn VPC, air-gapped, and IL6 deployment requirements.

Uri MerhavUpdated Mar 31, 2026 · 12 min

Insights

ISO 27001, SOC 2, and Compliance for AI Pipelines

AI wrappers fail vendor security assessments. Learn what FedRAMP, HIPAA, and AI governance standards require for compliance.

Uri MerhavUpdated Mar 30, 2026 · 14 min

Technical

Preventing AI Hallucinations with Visual Review & Provenance

LLM hallucinations destroy trust in document AI. Structure-preserving parsing and bounding-box provenance enable visual auditability.

Uri MerhavUpdated Mar 30, 2026 · 12 min

Technical

Processing Massive Documents: Splitting & Classification

Process 1,000-page concatenated PDFs by intelligently splitting them into logical documents and routing each to specialized extraction.

Nitai DeanUpdated Mar 30, 2026 · 12 min

Technical

Schema-Driven Governance: Forcing AI to Speak Your ERP's Language

Your LLM returns 'invoice_date' one time, 'date' the next. Your ERP crashes. Schema enforcement makes AI outputs match your database - every time.

Nitai DeanUpdated Mar 30, 2026 · 12 min

Customer Stories

Bardavon Health automates referral intake with DocuPipe

is a workers' compensation care coordination company that manages treatment for injured workers across the United States. With document extraction powered by DocuPipe, Bardavon standardizes referral intake packages, pulling structured data from medical records, prescriptions, and insurance documents to support case management.

Technical

Document Processing for Secure RAG Pipelines

Garbage in, hallucination out. RAG pipelines fail on flat text. Structure-preserving parsing provides clean, layout-aware evidence for enterprise LLMs.

Nitai DeanUpdated Mar 27, 2026 · 12 min

Technical

Handwriting & Checkboxes at Scale: Why Gov & Healthcare Run on Paper

Your OCR returns garbage on rushed handwriting and ambiguous checkboxes. Multimodal AI reads messy forms the way humans do - with context.

Uri MerhavUpdated Mar 27, 2026 · 12 min

Technical

The Human-in-the-Loop (HITL) Escalation Protocol

100% automation is a myth in high-risk environments. Confidence triggers and structured escalation protect regulated document processing.

DocuPipe TeamUpdated Mar 27, 2026 · 12 min

Technical

Table Extraction for Complex Government & Financial Filings

Merged cells become chaos. Multi-page tables lose headers. Your analyst spends hours on manual cleanup. Structure-preserving extraction fixes this.

Nitai DeanUpdated Mar 27, 2026 · 13 min

Insights

Total Cost of Processing (TCP) in Highly Regulated Industries

ROI calculators miss the real costs: audit failures, manual redaction, and regulatory penalties. The TCP Model uses BLS wage data for accurate economics.

Uri MerhavUpdated Mar 27, 2026 · 14 min

Insights

LLMs for Document Processing: What Works and What Doesn't

LLMs can extract document data without training, but they hallucinate without confidence scores. Learn what works and why modern IDP combines LLMs with traditional ML.

Uri MerhavUpdated Mar 26, 2026 · 8 min

Customer Stories

Political Financial Management automates donation compliance with DocuPipe

is a compliance and bookkeeping firm serving political campaigns across the United States with FEC reporting and donation management. With document extraction powered by DocuPipe, PFM transforms batch-scanned checks and direct-mail donation slips into structured data ready for federal compliance filings.

Customer Stories

Smile America processes 300,000+ handwritten forms with DocuPipe

is one of the largest school-based dental care providers in the United States, delivering preventive services to hundreds of thousands of children across 19 states. With handwriting recognition powered by DocuPipe, Smile America automates patient intake from 300,000+ paper consent forms annually.

Customer Stories

Domos built a unified document layer with DocuPipe

is a property management technology company building an AI workforce that automates routine tasks and communications for residential property managers and operators. With document intelligence powered by DocuPipe, Domos processes ACORD certificates, violation notices, and property documents across their platform.

Insights

Form Data Extraction: When Structured Documents Are Surprisingly Hard

Forms have fixed fields. Extraction should be easy. It's not. Checkboxes, handwriting, and version variations break most tools. Here's what actually works.

DocuPipe TeamUpdated Mar 25, 2026 · 8 min

Customer Stories

Huliot Group automates Hebrew purchase order processing with DocuPipe

is proud to be a global leader in advanced flow systems, designing, manufacturing & marketing complete solutions for water supply, wastewater & infrastructure. With automated document processing powered by DocuPipe, Huliot Group extracts and classifies Hebrew purchase orders across its companies, automatically feeding structured data into their Priority ERP and accelerating order fulfillment.

Insights

IDP vs OCR: Why OCR Alone Leaves You Stuck with Raw Text

OCR gives you text. IDP gives you data. If you're still writing regex to parse OCR output, you're solving the wrong problem.

Nitai DeanUpdated Mar 25, 2026 · 7 min

Insights

IDP vs RPA: Why Choosing One Over the Other Is a Mistake

Most teams pick RPA or IDP. Most teams get it wrong. Here's why the best document workflows use both, and what breaks when you don't.

Nitai DeanUpdated Mar 25, 2026 · 8 min

Insights

Document Extraction: The Boring AI Use Case That Actually Works

Forget chatbots. The GenAI use case with the clearest ROI is document extraction. Here's why LLMs cracked a problem that stumped engineers for decades.

Uri MerhavUpdated Mar 24, 2026 · 7 min

Technical

Can Public Mining Data Predict Stock Prices? Part 1: Getting the Data

Ontario publishes thousands of geological survey documents. Most investors ignore them. We're scraping them all to see if they predict mining stock prices.

Nitai DeanUpdated Mar 24, 2026 · 5 min

Technical

Web Scraping the Ontario Geological Survey

In this episode we wrap up the data scraping process for mining data in Ontario. With our trusty LLM sidekick, we build a script that uses Selenium to scrape data from the OGS website. We deal with issues like the data being too big, and how to properly iterate when the results are not what we expect.

Nitai DeanUpdated Mar 24, 2026 · 5 min

Technical

PDF Data Extraction: Why the Format That Preserves Everything Shares Nothing

PDFs were built to look perfect, not to share data. That 1993 design choice is why your team still copies and pastes. Here's how to actually get the data out.

Uri MerhavUpdated Mar 23, 2026 · 7 min

Insights

Vector Similarity Search is Hopeless

A pre-baked notion of similarity is inherently flawed and destined to fail. The only way to achieve reliable and accurate retrieval is through intentional document indexing tailored to unique business needs and constraints. This post explores a fundamental insight that drove us to build DocuPipe.

Uri MerhavUpdated Mar 23, 2026 · 7 min

Insights

What is Intelligent Document Processing? The Complete Guide for 2026

Intelligent document processing (IDP) is AI-powered software that converts documents into validated, structured data. Learn how IDP works, compare IDP vs OCR, explore use cases, and see the ROI.

Uri MerhavUpdated Mar 23, 2026 · 14 min

Company

DocuPipe is now SOC 2 Type II and ISO 27001 compliant

DocuPipe is now SOC 2 Type II and ISO 27001 compliant. No breaches were found. We used the audits to tighten reliability and privacy controls, including monitoring every load balancer. Done in partnership with Vanta.

Uri MerhavUpdated Mar 22, 2026 · 2 min

Insights

Invoice Data Extraction: Automate AP with AI (2026 Guide)

Invoice data extraction uses AI to pull vendor info, line items, and totals from invoices automatically. Learn how it works and what accuracy to expect.

Nitai DeanUpdated Mar 22, 2026 · 8 min

Insights

Building a SaaS that makes money in under 24 hours

Learn how I turned an idea into a functional, revenue-generating SaaS product in just 24 hours. This post isn't about billion-dollar startups; it's about rapid execution, validation, and creating value. Whether you're bootstrapping or exploring, discover how to dip your toe into entrepreneurship quickly and effectively.

Uri MerhavUpdated Mar 22, 2026 · 10 min

Insights

Bank Statement Data Extraction: Automate Financial Document Processing (2026)

Bank statement extraction pulls transactions, balances, and account info from statements for lending, accounting, and verification. Learn what accuracy to expect.

Uri MerhavUpdated Jan 26, 2026 · 8 min

Insights

Cloud vs On-Premise Document Extraction: How to Choose (2026)

Cloud vs on-premise document extraction: learn which deployment model fits your security, compliance, and performance requirements. Includes hybrid options.

Uri MerhavUpdated Jan 26, 2026 · 8 min

Technical

Table Extraction from Documents: The Complete Guide (2026)

Table extraction pulls structured data from complex tables in PDFs and scans. Learn why tables break most extraction tools and what actually works.

Nitai DeanUpdated Jan 26, 2026 · 8 min

Insights

Build vs Buy Document Processing: An Honest Guide (2026)

LLMs made document extraction prototypes easy. Production systems are another story. Here's an honest framework for the build vs buy decision.

Uri MerhavUpdated Jan 17, 2026 · 9 min

Insights

Healthcare Document Automation: A Complete Guide (2026)

Healthcare document automation uses IDP to process claims, patient records, and referrals - cutting admin time while staying HIPAA compliant.

DocuPipe TeamUpdated Jan 16, 2026 · 8 min

Insights

Why AI Fails to Understand Documents

You uploaded a PDF to ChatGPT and it answered your question. That means it read the whole document, right? Wrong. Here's what's actually happening under the hood, why it leads to spectacular failures, and how to build document processing that actually works.

Uri MerhavUpdated Dec 19, 2025 · 12 min

Company

DocuPanda is now DocuPipe.ai

As our service matures, so does our brand. DocuPipe.ai has a ring to it (we hope). It captures what we're all about: a document AI extraction tool that is extremely reliable and consistent. Read on to learn about our motivation for rebranding, effective May 15th, 2025.

Uri MerhavUpdated May 9, 2025 · 2 min

Company

Introducing DocuPipe's Integration with Make

Discover how DocuPipe's new integration with Make simplifies automation. With over 2,000 supported apps, connect your document workflows to tools like QuickBooks, HubSpot, and Google Sheets—no coding required. Streamline data extraction, enhance productivity, and build robust connections without spending engineering resources!

Uri MerhavUpdated Nov 7, 2024 · 2 min