Ishqkala uses a small number of third-party subprocessors to deliver the platform. This page lists every subprocessor that may process Customer Data, what each one does, where the data is located and which categories of data it can reach. Customer claim files, including bodily injury medical records, are stored and processed only in Amazon Web Services US regions.
How we choose and oversee subprocessors
Before onboarding a subprocessor, our security team reviews its SOC 2 Type II or ISO/IEC 27001 report, its encryption and access controls, and its data location commitments. Each subprocessor signs written data protection terms at least as protective as our Data Processing Addendum, and any subprocessor that could receive protected health information also signs a business associate agreement. Subprocessors are re-reviewed every year.
We send each subprocessor only the data its function requires. Only Amazon Web Services stores or processes claim files. The other services on this list receive account, contact or operational data, and claim content is redacted or blocked before it reaches them.
Model inference stays in our AWS accounts
Ishqkala's document understanding and liability reasoning models run on compute inside Ishqkala-controlled AWS accounts in US regions, in the same region as the customer's tenant. No third-party AI model provider receives claim files, adjuster notes, extracted fields or model outputs.
No customer data is used to train, fine-tune or evaluate any model, ours or anyone else's. Inference inputs and outputs are kept within the customer's tenant boundary and are subject to the same retention and deletion terms as the rest of Customer Data.
Data locations
Production runs in AWS US East (N. Virginia), with disaster recovery in AWS US West (Oregon). Customers who elect AWS GovCloud (US) are hosted entirely in GovCloud regions.
Our engineering and support teams in India are Ishqkala personnel, not a subprocessor. Their access to production systems is governed by our Data Processing Addendum and takes place only through virtual workspaces hosted in the US region where the data resides. No Customer Data is stored in India.
Services that never receive Customer Data
We use website analytics and a customer relationship management system for ishqkala.si and for business contacts. These tools are not connected to the platform, never receive claim data, and are covered by our Privacy Policy and Cookie Policy rather than this list.
Changes to this list
We give at least 30 days' notice before adding or replacing a subprocessor. Notices go by email to each customer's designated privacy and security contacts. To add a contact to the notice list, email privacy@ishqkala.si with the subject line "Subprocessor notices".
Customers may object to a new subprocessor on reasonable data protection grounds within the notice period. We will work with the customer to address the concern, and if we cannot, the customer may terminate the affected service as described in the Data Processing Addendum.
If a subprocessor must be replaced urgently to protect Customer Data, for example after a security incident at that vendor, we may make the change with shorter notice. We will tell affected customers as soon as possible, explain the reason, and keep the same objection rights open for 30 days after the change.
What every subprocessor contract requires
Each subprocessor agreement requires the vendor to process data only on Ishqkala's instructions, to keep it confidential, to encrypt it in transit and at rest, to notify us promptly of any security incident, to support our customers' audit and regulatory requests, and to delete or return data when the service ends. We remain responsible to our customers for each subprocessor's performance.
Customers who need a copy of a subprocessor's security report or our review notes for their own vendor oversight program can request them under NDA at security@ishqkala.si. Other questions about this list can go to privacy@ishqkala.si.
| Subprocessor | Purpose | Data location | Data categories |
|---|---|---|---|
| Amazon Web Services, Inc. | Cloud infrastructure: compute, storage, managed databases, encryption key management and hosting for Ishqkala model inference | United States: US East (N. Virginia) and US West (Oregon); AWS GovCloud (US) on customer election | All Customer Data, including claim files, adjuster notes, estimates, photographs and bodily injury medical records |
| Okta, Inc. (Auth0) | Platform user authentication, SAML and OpenID Connect single sign-on, and multi-factor authentication | United States | Customer user names, business email addresses, roles and authentication logs |
| Datadog, Inc. | Infrastructure monitoring, application performance tracing and security log analysis | United States | System metrics, service logs and user identifiers; claim content and medical information are redacted before export |
| Twilio Inc. (SendGrid) | Transactional email for SOL deadline alerts, sweep completion notices, user invitations and password resets | United States | Recipient names and business email addresses; alert text with masked claim references |
| Zendesk, Inc. | Customer support ticketing and help center | United States | Requester contact details and ticket text; claim file attachments are blocked and must be shared through the platform |
| Cloudflare, Inc. | DNS, DDoS protection and web application firewall for app.ishqkala.si, with TLS termination restricted to US data centers | United States | IP addresses, request headers and request metadata in transit |
| Google LLC (Google Workspace) | Business email, calendars and documents for customer correspondence and implementation planning | United States | Customer personnel contact details and correspondence; claim files are not stored in Workspace |