Trust

Security, Data Handling & IP Ownership

You own everything we build: 100% of the source code, workflows, prompts and documentation, transferred at handover with no vendor lock-in. Your data is never used to train AI models, AI systems are deployed privately, and every agent runs with least-privilege access, human approval for high-impact actions and full audit logs.

You own 100% of the code and IP

  • All source code, workflows, prompts, infrastructure configuration and documentation we create for you are transferred to you at handover.
  • Code lives in your repositories and runs in your cloud accounts wherever possible, never locked to ours.
  • Credentials, API keys and admin access are handed over and our access is removed at the end of the engagement unless you keep us on a support retainer.
  • No proprietary platform or licence fee sits between you and your system, so there is no vendor lock-in.

Your data stays yours

  • We use enterprise AI APIs (OpenAI, Anthropic, Azure OpenAI, AWS Bedrock) that do not train on your data by default, or self-hosted open-source models when data must never leave your infrastructure.
  • RAG indexes and vector databases are deployed in a private environment in your cloud or a dedicated environment we manage for you.
  • We work with the minimum data needed to build and test, and use anonymised or synthetic samples where possible.
  • Project data on our side is deleted at the end of the engagement on request.

How we secure what we build

  • Encryption in transit (TLS) and at rest for databases, storage and backups.
  • Least-privilege access: every service, agent and integration gets only the permissions it needs.
  • AI agents run with scoped tool permissions, spend limits, human approval for high-impact actions and full audit logs.
  • Inputs and outputs of LLM features are validated to mitigate prompt injection and data leakage, following the OWASP Top 10 for LLM Applications.
  • Secrets are kept in a secrets manager, never in code or workflow exports.

Regulated industries

  • For healthcare projects we design for HIPAA compliance and deploy only on infrastructure and AI providers that sign a Business Associate Agreement (BAA).
  • For SOC 2 and GDPR requirements we document data flows, access controls and retention so your compliance team can review them.
  • Compliance depends on how a system is operated as well as how it is built, so we hand over the documentation your auditors will ask for.

Confidentiality

  • We are happy to sign your NDA before a scoping call or discovery.
  • We never publish client names, logos, screenshots or results without written permission.

Frequently asked questions

Who owns the code Lytnex writes?

You do. All source code, workflows, prompts, configuration and documentation created for your project are transferred to you at handover, along with credentials and admin access.

Will our data be used to train AI models?

No. We use enterprise AI APIs that do not train on your data by default, or self-hosted open-source models when data must stay on your own infrastructure.

Will you sign an NDA?

Yes. We are happy to sign your NDA before a scoping call or discovery.

Can you build HIPAA-compliant AI systems?

We design healthcare systems for HIPAA compliance, with encryption, access controls and audit logging, and deploy only on infrastructure and AI providers that sign a Business Associate Agreement.

See also our Terms of Service and Privacy Policy.

Have a security questionnaire?

Send it with your booking and we'll walk through it on the call.

Book a Strategy Call