Privacy Policy

1. Who we are

Parity Layer ("Parity Layer", "we", "us") provides a service that proves whether a cheaper AI model matches or beats a customer's current model on their own prompts, and routes traffic to proven models. This policy explains what personal data we collect, why, and the choices you have. It applies to paritylayer.com, our dashboard, and our API.

Contact for anything in this policy: roman@paritylayer.com.

2. What we collect

Account data: name, email address, and authentication identifiers, handled through our sign-in provider (Clerk).

Billing data: payment details are processed by Stripe. We do not store full card numbers on our systems.

API traffic content: when you route requests through Parity Layer, we process the prompts and responses in those requests. This is the core of the service: we run candidate models against your real traffic and compare outputs to produce quality proofs. Comparison content is retained so that proofs remain auditable.

Uploaded exports: if you use the offline proof, we process the request log file (for example a JSONL export) you upload for the purpose of producing your proof.

Usage and analytics data: product usage events and website analytics (PostHog), plus standard server logs for security and reliability.

Communications: emails you exchange with us, including support and onboarding messages.

3. How we use data

To provide the service: forwarding requests to your chosen model provider, running parallel candidate models, judging output quality, producing savings and proof reports, and billing.

To secure and improve the service: debugging, abuse prevention, capacity planning, and product analytics.

To communicate with you: service emails (proofs ready, approvals, billing) and, where you have agreed, product updates.

We do not sell personal data. We do not use your prompts or outputs to train our own foundation models.

4. Subprocessors

We rely on a small set of infrastructure providers to run the service: Clerk (authentication), Stripe (payments), Railway (hosting), Neon (database), PostHog (analytics), Resend (transactional email), and the AI model providers you configure or that we route to on your behalf (for example OpenAI, Anthropic, Google, xAI, Groq, Together AI). Each receives only the data needed for its function.

5. Retention

Account and billing records are kept while your account is active and as required for legal and accounting purposes.

Comparison content and proof records are retained so that historical proofs remain verifiable. If you want specific content deleted, contact us and we will action it unless we are legally required to keep it.

You can request deletion of your account and associated personal data at any time.

6. Security

Data is encrypted in transit. Access to production systems is restricted and credentialed. Provider API keys you supply are stored encrypted and used only to route and prove your own traffic.

7. Your rights

Depending on where you live (including under UK and EU GDPR), you may have rights to access, correct, export, restrict, or delete your personal data, and to object to certain processing. Email roman@paritylayer.com and we will respond within the statutory period. You may also complain to your supervisory authority (in the UK, the ICO).

8. International transfers

Our infrastructure providers may process data outside your country, including in the United States. Where required, transfers rely on appropriate safeguards such as standard contractual clauses.

9. Changes

We will post any changes to this policy on this page and update the date below. Material changes will be notified by email to account holders.