Trust & Security

mia is built to earn trust.

A simple security model: your data stays in your environment, you control the AI, and every insight can be traced back to its source.

Your tenant, your data

mia installs inside your Microsoft Azure tenant. Data is stored on your subscription, billed to your account.

We do not train on your data

Your prompts and documents are never used to train any AI model. Ever. They stay inside your environment.

Every claim has a source

mia links every insight back to its evidence. Unsupported claims are removed before they reach your team.

Works with your identity provider

Microsoft Entra ID, Okta, or any SAML 2.0 provider. SCIM provisioning and role-based access included.

Our commitments

Six guarantees, not optional extras.

These apply to every mia deployment, whether you are a mid-market team or a global enterprise.

Tenant-native by default

Runs in your Azure tenant, not ours. Storage, compute, and network are under your control.

Customer-controlled AI

Pick the model you trust. Your data is never used for training or evaluation.

Evidence over assertion

Every output shows the source. Unsupported claims are filtered out before delivery.

Enterprise identity

Single sign-on, automatic user provisioning, and fine-grained permissions by default.

Full audit trail

Logs for every login, export, and configuration change. Streamed to your SIEM if needed.

Built to stay online

Stateless services, backups, tested recovery. Built for teams that cannot afford downtime.

Where your data lives.

Your tenant. Your region. Your keys. We do not move your data outside the environment you control.

European by default

Microsoft Azure West Europe (Amsterdam) or North Europe (Dublin). Optional OVHcloud regions in France.

Customer-managed keys

Encrypt data with your own keys in Azure Key Vault. Rotate or revoke keys on your schedule.

No shared environments

Each customer has a separate environment. No shared indexes, caches, or memory between tenants.

Clear sub-processor list

Small, named, GDPR-bound list. You are notified of changes and can decline a sub-processor.

Plain language

mia does not combine, anonymize, or share your data with other customers. If you leave, your environment is yours to delete.

Trustworthy outputs, by design.

A decision intelligence tool is only useful if you can trust what it tells you. mia checks every claim before it reaches you.

1

Retrieve sources

mia gathers sources and scores them on freshness, authority, and relevance.

2

Check every claim

A reviewer checks each claim against the sources. Unsupported or speculative claims are flagged.

3

Fix before delivery

Flagged claims are rewritten, removed, or re-anchored. Only supported insights reach your team.

No model lock-in

Use Azure OpenAI, Microsoft Phi, Anthropic on Azure, or self-hosted open models. Switching is a configuration change.

Aligned with AI governance standards

We follow the NIST AI Risk Management Framework and ISO/IEC 42001 principles. EU AI Act developments are tracked and shared with customers.

Production posture

Infrastructure that stays reliable.

Built on Microsoft Azure as containerized microservices. Scalable, recoverable, and ready for customer-controlled deployment.

Runs on Microsoft Azure

AKS with autoscaling, multi-zone replicas, and managed updates.

Stateless services

Each service is independent. A failure in one does not affect the rest.

GitOps deployments

Every change is reviewed, signed, and reversible through pull-based GitOps.

Backups and recovery

Continuous backups plus 12-hour snapshots, with recovery tested quarterly.

Compliance status.

We only list certifications we hold or are actively completing. Reports are available on request under NDA.

FrameworkStatusNotes
ISO 27001In progressISMS scoped, controls implemented, audit scheduled
GDPROperationalDPA, sub-processor list, and DPIA template available
Microsoft Azure inheritanceNativeTenant deployments inherit Azure ISO 27001, HIPAA, FedRAMP posture

Common questions.

Quick answers for procurement, security, and legal teams.

Strategic partners

Built with partners that share our standards.

Our infrastructure and ecosystem partners help us deliver secure, sovereign, and responsible AI deployments.

Partner

Microsoft for Startups

Access to Azure infrastructure, Azure OpenAI Service, and Microsoft engineering and go-to-market support.

  • Azure-native architecture
  • Azure OpenAI and AI safety frameworks
  • Enterprise go-to-market support

Partner

OVHcloud

Sovereign European cloud option for strict data residency or SecNumCloud-aligned requirements.

  • European data residency guarantees
  • Trusted Cloud and SecNumCloud-aligned infrastructure
  • Carbon-efficient data centers

Get the security pack for your CISO.

Architecture diagram, sub-processor list, DPA, DPIA template, and pen-test summary. One PDF, NDA optional.

Book a demo