Skip to content
Version 0.1.0Effective 2026-09-06

Privacy Policy

This policy explains what Hiro Labz collects when you use HiroAI, why, who else sees it, how long it is kept, and the choices you have.

Draft for legal review

This document has not been approved by Legal. It shows what the product does today so the review can start from the truth, and it is not yet a binding agreement.

1. Who this covers

This policy covers people who create or use a HiroAI account (workspace owners and members) and, where your agents talk to your own customers, the end users whose conversations pass through the service. For end users, you (the workspace) decide what is collected and why; we process it on your instructions.

2. What we collect

We collect the following categories of data:

  • Account data: your name, email address, password (stored hashed), company details you enter, and sign-in events. If you sign in with a social account, the identity provider shares your name, email and an identifier.
  • Workspace content: the prompts, instructions and settings you write for your agents; the documents you upload to knowledge bases and the text extracted from them; the conversations your agents have; the memory facts agents store about your customers; and the traces of each agent run, including the inputs and outputs of each step.
  • Usage and billing data: which features you use, credits consumed per agent and model, plan and payment status. Card details are handled by our payment processor and never reach our servers.
  • Technical data: IP address, browser and device information, and the request logs needed to run and secure the service.
  • Product analytics: anonymised interaction events and masked session recordings, described in section 5.

3. How we use it

We use this data to:

  • Provide the service: run your agents, retrieve from your documents, show you traces and reports, and meter credits.
  • Keep the service secure: detect abuse, enforce rate limits and budgets, and investigate incidents.
  • Support you: answer your questions and diagnose problems you report.
  • Improve the product: understand which features are used and where people get stuck, using the masked analytics described in section 5.
  • Meet legal obligations, including tax and accounting records.
  • We do not sell your data, and we do not use your workspace content to train our own models.

4. Who else processes it

Third-party model providers you select. When an agent runs, the prompt, the retrieved content and the conversation are sent to the model provider you chose in your workspace so it can generate a response. We send nothing to a provider you have not selected. Each provider's own privacy terms apply to what it does with that data.

Tools and channels you connect. If you connect an external system (for example a CRM or a webhook endpoint), your agents send data to it on your instructions.

Our own service providers. We use hosting, database, email, payment and analytics providers to run the service. They process data only on our instructions. We do not publish a formal list of these providers at this time; contact us if you need it for a procurement review.

Authorities. We disclose data where the law requires it.

5. Analytics and session recording

We use PostHog for product analytics, error reporting and session recording in the console. Recordings help us see where the interface confuses people. Because the console shows your prompts, documents and conversations, recording is masked by default:

  • Everything you type into any field is masked in recordings. We see that a field was filled, not what it contained.
  • Chat transcripts, prompt editors, extracted document text and trace payloads are masked as whole regions. Their text is replaced before a recording leaves your browser.
  • Clicks inside those regions are recorded without the text of the element you clicked.
  • Browser console output is not collected. Error reports are collected so we can fix crashes; an error message may occasionally include a fragment of application state, and we treat those reports as confidential.
  • Analytics events carry identifiers and timings (for example which onboarding step completed and how long it took), never prompt or document content.

6. Retention and deletion

Workspace content (prompts, documents, conversations, memory and traces) is kept until the workspace deletes it or closes the account. You can delete individual documents and conversations, wipe workspace memory from Settings, and delete agents with their runs.

Test mode keeps test agents, keys and data separate from live ones so you can experiment without mixing the two. It is a data boundary: test data is stored and retained in the same way as live data, and the same deletion controls apply.

Account and billing records are kept for as long as your account exists and afterwards for as long as tax and accounting law requires, then deleted or anonymised.

Technical logs are kept for a short rolling period for security and debugging. Analytics data is kept under our analytics provider's retention settings and is not linked back to workspace content.

7. Security

Data is encrypted in transit. Provider keys you store with us are encrypted at rest and are write-only: once saved, they cannot be read back through the console or API. Each workspace's data is isolated at the database layer so that one workspace cannot read another's records. Access by our staff is limited to what running and supporting the service requires.

No hosted service can promise perfect security, and we do not hold an independent security certification at this time. If you believe your account has been compromised, contact us straight away.

8. Your choices and rights

You can view and update your account details in Settings, export or delete your workspace content through the console and API, and close your account. Depending on where you live you may also have legal rights to access, correct, delete or restrict the use of your personal data, to object to processing, and to complain to a supervisory authority. To exercise them, contact us at the address below. If you are an end user of one of our customers' agents, please contact that customer first; we will help them respond.

Our lawful bases for processing, and the details of international data transfers, are [to be completed by Legal].

9. Children

The service is not directed at children and we do not knowingly collect account data from anyone under 18. If you believe a child has created an account, contact us and we will remove it.

10. Changes to this policy

We may update this policy. The version number and effective date at the top tell you which version applies. For material changes we will give notice in the console or by email before they take effect.

11. Contact

Privacy questions and requests: [contact email to be completed by Legal].