Deployed privately. Traceable to source.
(Re)insurance runs on regulated data. mea runs on your own single-tenant instance, integrates through the systems you already use, and traces every value back to the document it came from.
Security review stops being the thing that slows the deal.
- Single-tenant AWS deployment
- Regional data residency
- Private learning environment
- Field-level traceability
02 · Private Learning Architecture
Insurance-native from day one. Private learning by design.
mea arrives pre-trained on the language, documents and workflows of insurance. It then adapts to how your organization works, inside your own instance, with data that stays private to you.
Pre-trained for insurance
mea already understands insurance language, documents, entities and workflows before it reaches you. Nothing starts from zero.
Adapted to your guidelines
It takes on your terminology, schemas, operating procedures, appetite rules and workflows; all specific to how your business runs.
Improved through governed feedback
Corrections and approvals from your team refine your configuration, with every change reviewed and versioned.
Learning stays private
Each client operates on its own instance. Your data is your data, and the learning derived from it is private to you.
03 · Deployment and Integration
Rapid full deployment, alongside the systems you already run.
Rapid full deployment
Single-tenant deployment
No core replacement
Single-tenant AWS deployment
Your instance is yours. No shared processing tenancy, and dedicated resources configured for your workloads.
Regional data residency
Data handling and retention configured to the jurisdiction you operate in. Your data stays where your regulator expects it.
Direct integration
REST API, mailbox integration, SFTP and callbacks, with structured output in JSON, XML or Excel. mea Operations runs alongside policy administration, claims, rating, placing and finance systems.
Your identity provider
SSO through your existing identity provider. Role-based access matched to your permissions model, provisioned automatically.
05 · Certifications
Certified, and kept that way.
ISO 27001 Certified
GDPR Compliant
SOC 2 Compliant
NYDFS Compliant
HIPAA Compliant
FAQs
Does mea use our data to train models for other clients?
No. Each client operates on its own instance. Corrections and approvals from your team improve your configuration: your mappings, your rules, your workflows. Your data is your data. The insurance knowledge the platform arrives with comes from mea's own pre-training on insurance documents and workflows.
Is mea a black box?
No. Every output is supported by transparent reasoning and field-level lineage back to the source document. A decision made today can be explained and defended later, which is the standard regulated insurers are held to.
How does mea sharpen on our data over time?
mea Ora arrives pre-trained on insurance language, documents and workflows. From there it sharpens on your operation: your document formats, your terminology, your appetite rules and the confidence thresholds you set. Approved corrections feed back through a reviewed and versioned release process, so accuracy improves on the work you actually do.
What happens to throughput during renewal peaks and CAT events?
mea Operations runs serverless, scaling with demand rather than provisioned capacity, which is what allows the platform to absorb volume spikes without backlog and without a corresponding increase in headcount. Service commitments covering throughput, accuracy and uptime sit in the agreement.
Security by design
mea was built by former insurance operators and technology leaders from inside the industry, working alongside some of the world’s leading technology talent. The people who designed this platform have sat on the other side of enterprise security reviews, run the programs and carried the obligations. That is why traceability, access control and data isolation are in the architecture rather than added afterward.
Cut the drag. Just win.