Legal
Subprocessors
A list of third-party providers JestBot relies on to deliver the product, categorized by function, along with what data each one can access. Relevant to GDPR-conscious customers evaluating our data flows as part of a data processing agreement (DPA).
Last updated: Placeholder — set a real date on publish
1. Overview
JestBot uses a small, deliberately limited number of third-party providers to run the platform. Some process data for every customer by necessity — hosting, payments, authentication; others, specifically the BYOK model providers, only ever see your data if you personally choose to configure that provider’s API key on a bot. If you stick with the default Ollama-backed model, none of the BYOK providers listed below receive any of your data, since Ollama runs on our own self-hosted infrastructure rather than calling out to a third party.
2. What is a subprocessor
A subprocessor is any third party we engage that processes personal data on our behalf in order to deliver JestBot to you. This is distinct from a service you connect yourself for your own purposes outside JestBot’s core function — we’re listing providers that are structurally part of how JestBot itself operates.
3. Infrastructure & hosting
| Provider | Purpose | Data accessed |
|---|---|---|
| Hosting provider | Application hosting, database, file/document storage, vector embedding storage | All account, bot, document, embedding, and conversation data |
This is the foundational layer — every account, bot, document, embedding, and conversation record is stored here. See our Security page for how this data is encrypted and access-controlled.
4. Payments
| Provider | Purpose | Data accessed |
|---|---|---|
| Razorpay | Payment processing for INR and USD billing, subscription management | Billing name, email, payment method details (card/UPI/etc. handled directly by Razorpay, not stored by JestBot) |
JestBot does not directly store your full card number or payment credentials — those are handled entirely within Razorpay’s PCI-compliant systems. We store only the billing metadata needed to manage your subscription (plan, status, renewal date, transaction history).
5. Authentication & messaging
| Provider | Purpose | Data accessed |
|---|---|---|
| Google Firebase Authentication | Google sign-in (OAuth), OTP delivery for phone/email verification | Email address, phone number (for OTP), authentication identifiers and tokens |
| Firebase Cloud Messaging | Real-time push notifications to agents for handover requests and assignments | Device push token, notification content (e.g. "New handover request") |
6. BYOK model providers (opt-in only)
These providers only ever process data for bots where you’ve explicitly added that provider’s API key yourself. Each request made through your key is subject to that provider’s own data handling, retention, and training-use terms — JestBot passes your message content to them at query time but does not control what that provider does with it afterward, so we’d encourage reviewing each provider’s own policy if this matters for your use case.
| Provider | Purpose | Data accessed |
|---|---|---|
| OpenAI | Chat/embedding generation — only if you configure an OpenAI key on a bot | Visitor message content sent to the model at query time; not JestBot's other stored data |
| Anthropic | Chat generation — only if you configure an Anthropic key on a bot | Visitor message content sent to the model at query time |
| Google Gemini | Chat/embedding generation — only if you configure a Gemini key on a bot | Visitor message content sent to the model at query time |
| Groq | Chat generation — only if you configure a Groq key on a bot | Visitor message content sent to the model at query time |
| Mistral | Chat generation — only if you configure a Mistral key on a bot | Visitor message content sent to the model at query time |
7. How we vet subprocessors
Before adding a new subprocessor, we consider its own security practices, data handling terms, and whether it's necessary for a feature we're building, rather than adding third parties by default. We aim to keep this list as short as practically possible.
8. Data location
[Placeholder — state the actual regions/data centers used by your hosting provider here, e.g. "Primary data storage is located in [region]." This matters for customers with data-residency requirements and should reflect your real infrastructure setup, not be left generic.]
9. Updates to this list
We’ll update this page whenever we add or remove a subprocessor. For material changes affecting how your data is processed — for example, adding a new BYOK provider or changing hosting infrastructure — we’ll also notify account holders directly by email where required by applicable law, giving you the opportunity to object before the change takes effect where such a right applies under your agreement with us.
10. Questions for your DPA
If your organization requires a signed Data Processing Agreement referencing this subprocessor list, or has additional vendor security questionnaire requirements, reach out via the contact page.