Process ACORD forms, applications, loss runs, and supplemental documents from any broker or carrier. AI reads every format without templates or setup.
“Cut submission triage from 25 minutes to under 2 minutes. Our underwriters spend time evaluating risk now, not retyping data.”
“Processed 1,200 renewal submissions in a three-week window with no temp staff. Last year that same volume required six people working overtime.”
“Data accuracy on ACORD forms went from 89% with our old OCR vendor to 97.5% on the first pass. We barely touch the exception queue anymore.”
Audited controls over a sustained period, not a point-in-time check.
Bank-grade encryption at rest and TLS 1.2+ in transit.
Documents deleted within 24 hours. No copies retained.
Drag and drop submission packages, connect a cloud drive, or set up email auto-forwarding from your broker inbox. PDF, JPEG, PNG, TIFF, and digital documents all work.
The AI identifies ACORD form types, reads named insureds, coverage limits, loss history, and supplemental data by context—not fixed coordinates. Every broker format works on the first upload.
Get structured output in Excel, Google Sheets, CSV, or JSON. Use the REST API to push extracted data directly into your underwriting workbench or rating engine.
Last updated: June 2026
A commercial lines underwriter typically handles 40 to 60 submissions per week, each containing a combination of ACORD applications, loss runs, financial statements, supplemental questionnaires, and broker narratives. These documents arrive in varying formats from different brokers, and the underwriter’s initial task is always identical: review every page, locate the critical data points, and enter them into a workbench or spreadsheet. That manual triage absorbs 30 to 45 percent of an underwriter’s day before any risk evaluation takes place. For teams comparing platforms, our analysis of the best insurance underwriting software evaluates 8 solutions side by side. Insurance OCR tools have advanced considerably over the past decade, yet most still demand per-broker templates or carrier-specific configuration that breaks whenever a broker modifies their submission format.
AI-powered underwriting software interprets submission documents contextually. Rather than searching for data at fixed page coordinates, the AI recognizes that a named insured field on an ACORD 125 from Marsh contains the same category of information as a freeform cover letter from a regional broker—and extracts both accurately. This removes the template maintenance burden that hampered earlier OCR tools and ensures new broker relationships require no IT involvement. Our explainer on what underwriting automation is covers the foundational principles. Automated underwriting workflows built on this extraction layer can direct submissions to the appropriate underwriter, pre-fill workbenches, and identify incomplete packages before anyone opens the file. For concrete implementation approaches, see our guide on underwriting workflow automation.
Accuracy is most consequential on ACORD forms with handwritten entries and on scanned loss runs where image quality fluctuates. Lido delivers 97-98% accuracy on printed ACORD fields and 93-96% on handwritten annotations, with field-level confidence scores that enable underwriting teams to define their own review thresholds. For underwriting OCR specifically, the combination of strong first-pass accuracy and transparent confidence scoring gives underwriters sufficient trust in the extracted data to act without re-reading every source document.
Carriers and MGAs evaluating underwriting software should benchmark against their real submission volume: the most cluttered broker packages, the poorest quality scanned loss runs, and the ACORD forms with handwritten margin annotations. That production-grade test reveals more than any vendor demonstration. Insurance claims OCR follows the same principle—accuracy on clean documents is baseline; the true differentiator is performance on the documents that currently demand manual handling.
Lido processes ACORD applications (125, 126, 130, 140), supplemental questionnaires, loss runs, financial statements, broker submissions, and declarations pages. The AI identifies the document type and extracts relevant fields automatically, regardless of broker or carrier format.
AI-powered extraction reads each document contextually, identifying fields by meaning rather than fixed position. A submission package from one broker may include ACORD forms, while another sends freeform narratives and spreadsheets. Both are processed without separate templates or per-broker configuration.
Yes. The REST API returns structured JSON with field-level confidence scores, which can be routed to any underwriting workbench, policy administration system, or rating engine. Teams use the API to populate Guidewire, Duck Creek, Majesco, and custom-built workbenches directly from extracted submission data.
On clean handwritten entries like agent signatures, policy numbers written in margins, and handwritten notes on ACORD forms, the AI achieves 93-96% character accuracy. For printed and typed text on scanned submissions, accuracy exceeds 98%. Every extraction includes field-level confidence scores so your team can flag low-confidence fields for review.
The platform is SOC 2 Type 2 certified with annual audits. All data is encrypted with AES-256 at rest and TLS 1.2+ in transit. Documents are deleted within 24 hours of processing, and no copies are retained. BAA is available for organizations subject to HIPAA.
Start free with 50 pages. Upgrade when you’re ready.
Built on Lido’s OCR engine
Built on Lido’s OCR engine
Built on Lido’s OCR engine