What was the impact of the AI solution?

Claims processing time decreased from 30-60 minutes per claim to near real-time. Automation reduced human error and improved accuracy in claims processing. Customer satisfaction improved and the approach was expanded to other business workflows.

Source

GCP

Loadsure: near real-time insurance claim verification using Document AI and Gemini

Use case typeIntelligent document processingUpdated Jun 13, 2026

Loadsure, a global insurtech firm based in London, built an AI-powered claims verification system to automate extraction and verification from bills of lading, invoices, and shipping documents. The solution uses Google Cloud Document AI for OCR, document classification, and per-document-type extractors, and Gemini for cases requiring deeper contextual comprehension. The company reports more than 95% extraction confidence and has expanded the approach to other insurance workflows such as underwriting and pricing.

Organization: Loadsure
Industry: Insurance
Location: United Kingdom
Published: November 2024

Reported outcomes

+95%

extraction confidenceOther quantified impact

30 minutesprocessing time per claim60 minutesprocessing time per claim

Strategic outcomes

Speed & agilityEnabled near real-time claims processingCustomer experience & trustImproved customer satisfactionRisk & complianceReduced human error in claims processing

Primary read

Use case focus

Showing 3 of 3

1Document Processing
2Claims Automation
3Information Extraction

Manual verification of claims submission documents caused 30-60 minute processing times, delayed claim settlements, and introduced risk of human error and disputes.

Using Google Cloud's Document AI, Loadsure automated data extraction from claim documents and used different document extractors per document type.
The team used the labeling tool and pre-trained processors to improve model quality and iterated with generative AI models for workflows where contextual understanding mattered.
Gemini was used for more complex document processing and entity extraction cases with variable inputs.

Technologies

Document AI Gemini

Claims processing time decreased from 30-60 minutes per claim to near real-time.
Automation reduced human error and improved accuracy in claims processing.
Customer satisfaction improved and the approach was expanded to other business workflows.

Architecture

Loadsure built an AI-powered claims verification workflow on Google Cloud using Document AI to extract and classify claim documents such as bills of lading, invoices, and shipping documents, with different extractors per document type. The team iterated with Document AI processors and labeling tools, and used Gemini for cases requiring more contextual document understanding and entity extraction.

Sources & evidence1

Groundedness: 5/5Type: Blog PostPublished: Nov 5, 2024Publisher: Google CloudEvidence: VendorConfidence: Medium

Primary source

AI-generated summary. Verify important details with the linked sources before relying on this case.

Explore related AI use cases

Customer evidence