Security

Your books hold some of the most sensitive data in your business. Here is how GeekyBooks protects it, from the infrastructure up to the way our AI agents handle your records.

Last updated: 23 October 2026

Our approach

Security is built into how GeekyBooks is designed, developed and operated. We follow a defence-in-depth approach: several independent layers of protection, so that no single failure exposes your data.

This page gives an overview of our current practices. It will be updated as our programme grows.

Security built on trusted cloud infrastructure

GeekyBooks is built on Microsoft Azure, one of the most secure and widely trusted cloud platforms. We use Azure’s enterprise-grade infrastructure, security controls and compliance capabilities to help protect your financial data.

How your data is protected

Industry certifications
Azure maintains a broad portfolio of industry and regional compliance certifications and standards.
Tenant isolation
Azure keeps GeekyBooks separate from other Azure customers and each GeekyBooks customer’s data sits in its own partition.
Physical security
Azure data centres are protected by multiple layers of physical security, controlled access and round-the-clock surveillance.
Threat protection
Azure’s platform includes built-in protection against network attacks, including distributed denial-of-service (DDoS).

Data encryption

  • In transit: all traffic between your browser or the mobile app and GeekyBooks is encrypted using TLS and data stays protected as it moves across the platform.
  • At rest: customer data is encrypted at rest within Microsoft Azure, providing an additional layer of protection for sensitive financial information.
  • Secrets: credentials, API keys and connection tokens are kept in a dedicated secrets store and are never written to application logs.

Access control

  • Roles and permissions: workspace owners decide who can view, create, approve or export data, down to individual modules.
  • Approvals: sensitive actions such as bill payments can require one or more approvers before they take effect.
  • Internal access: GeekyBooks staff access to production systems is restricted to a small number of authorised people, on a need-to-know basis and is logged.

Infrastructure

GeekyBooks runs on managed Microsoft Azure infrastructure, in physically secured data centres and benefits from Microsoft’s underlying platform security, infrastructure management, monitoring and protection capabilities. Production environments are isolated from development and testing and network access is restricted with firewalls and private networking.

Compliance: Microsoft Azure maintains a broad portfolio of industry and regional compliance certifications and standards. These cover the Azure infrastructure GeekyBooks is built on; GeekyBooks remains responsible for the security of its own application and the way it handles your data.

Tenant isolation: Azure’s multitenant architecture uses logical isolation to keep GeekyBooks workloads and data separated from those of other Azure customers. Within GeekyBooks, every customer is a unique tenant and each customer’s data is securely siloed in separate data partitions.

Monitoring and audit logs

We monitor our systems continuously for availability, errors and suspicious activity, alongside Azure’s layered platform security and monitoring. Within your workspace, key actions (such as edits to bills, approvals and changes to user access) are recorded in an audit trail so you can see who did what and when.

Breach notification: if we become aware of a security breach that affects your data, we will notify affected customers without undue delay, explain what happened and the steps we are taking and inform regulators where the law requires.

Backups and continuity

Customer data is backed up regularly to separate, encrypted storage. We test restoration procedures so that we can recover from hardware failure or other incidents with minimal disruption.

AI agents and your data

  • AI agents act only inside your workspace and within the permissions you set.
  • Agent suggestions, such as bill coding, PDF to CSV conversion, reconciliation or transaction matches, are visible and reviewable and you stay in control of what is posted to your books.
  • Where we use third-party AI model providers, data is shared only as needed to deliver the feature you are using.