Normalized claim
Time: 10 minutes decrease
Policy document processing time dropped from 10 minutes to under 60 seconds.
OneAssure: agentic insurance document processing with Document AI + Gemini Enterprise Agent Platform
Use case typeCustomer service automationUpdated Jun 13, 2026
OneAssure is a Bengaluru insurtech platform in India that simplified health insurance workflows using Google Cloud AI. The company used Document AI to extract structured information from unstructured policy PDFs and deployed the solution on Google Kubernetes Engine. It also used Gemini Enterprise Agent Platform to support product and policy Q&A, coverage gap analysis, and a Gemini-powered D2C chatbot for real-time policy assessments.
Reported outcomes
+90%
quantified impactCustomer experience
10 minutestime60 secondstime95-98%accuracy30%quantified impact+20%quantified impact
Strategic outcomes
New product / capabilityAutomated insurance document extractionSpeed & agilityStreamlined policy document processingCustomer experience & trustAdded real-time policy advisory chatbotBetter decisions & insightImproved policy assessment accuracy
Why do we believe this?Outcome claims, sources, and evidence checks
Normalized claim
Time: 60 seconds decrease
Policy document processing time dropped from 10 minutes to under 60 seconds.
Normalized claim
Accuracy: 95-98% increase
Extraction accuracy improved from 95% to 98%.
Normalized claim
Quantified impact: 30%
Manual query volume between teams and partners fell by 30%.
Normalized claim
Quantified impact: 20% increase
The AI advisory tools contributed to a 20% increase in active users and helped maintain about 90% advisor retention.
Normalized claim
Quantified impact: 90% increase
The AI advisory tools contributed to a 20% increase in active users and helped maintain about 90% advisor retention.
Why do we believe this deployment?Customer identity, provider attribution, maturity, and source checks
- Customer
- OneAssure
- Provider
- GCP
- Maturity
- Scaled Production
- Linked source
- Google Cloud Customer Story
As OneAssure scaled to about 1,000 new users per month, manual query volume and advisor workload increased
Customer identity supportedSource describes one deploymentMaturity supported
Primary read
Use case focus
Showing 3 of 5
- 1Document processing
- 2Intelligent document extraction
- 3Customer service automation
- Manual processing of complex health insurance PDFs required staff to type data into up to 30 fields per document.
- The process took about 10 minutes per file and was prone to human error, creating payout inaccuracies and disputes.
- As OneAssure scaled to about 1,000 new users per month, manual query volume and advisor workload increased.
- OneAssure built an AI extraction tool with Document AI for structured data extraction from insurance documents.
- The team deployed the solution on Google Kubernetes Engine to automate processing.
- Using Gemini Enterprise Agent Platform, OneAssure tuned model selection and reasoning settings to balance cost, latency, and accuracy.
- A Gemini-powered chatbot was added to provide real-time policy assessments and coverage gap analysis for customers and advisors.
- Policy document processing time dropped from 10 minutes to under 60 seconds.
- Extraction accuracy improved from 95% to 98%.
- Manual query volume between teams and partners fell by 30%.
- The chatbot generated 150 to 200 high-quality monthly leads.
- The AI advisory tools contributed to a 20% increase in active users and helped maintain about 90% advisor retention.
Architecture
OneAssure deployed an AI extraction workflow on Google Kubernetes Engine using Document AI for document parsing and Gemini Enterprise Agent Platform for model selection, reasoning optimization, and agentic Q&A. The article also describes a Gemini-powered chatbot for customer advisory interactions.
Sources & evidence2
Evidence: High75/100Evidence strength
- Customer explicitly identified
- Deployment status explicitly supported
- Primary source available
- Quantified outcome available
- Technical implementation details available
- Multiple corroborating sources available
Type: Customer StoryPublished: May 13, 2026Publisher: Google Cloud Customer StoryEvidence: PrimaryConfidence: High
AI-generated summary. Verify important details with the linked sources before relying on this case.
Explore related AI use cases
Customer evidence
Provider evidence
AI patterns
Browse the catalog
Was this useful?
Community
Comments
No published comments yet.