Said in the terms your own security team uses. What is true today, and what is not yet.
Xolarc works from the knowledge your enterprise has built over decades, and it touches customer data only for the specific requirement a person has put to it. It does not wander your systems to take information in. The controls enterprise teams expect are included, such as single sign-on, role-based access, audit logs, tenant isolation and data lifecycle management.
We do not access or integrate with any of your customer systems. Customer data enters Xolarc only when your firm exports and uploads it for a specific requirement.
Customer data is accessible only to people authorised on that account — no member of another account can see it, and access is granted by a person, never assumed. Your firm’s internal knowledge remains shared.
Single sign-on and directory provisioning in enterprise deployments; MFA on non-federated accounts. Five permission levels, least privilege by default.
One partition per enterprise, enforced at the database: row security on every table, a search index per tenant, working files scoped to the engagement.
AES-256 at rest, TLS 1.2+ in transit, stored credentials double-encrypted. Personal data is screened before storage or any model call; every detection is logged.
Model access runs on contracted terms that exclude training, or inside your own cloud subscription. Generated code executes in an isolated, ephemeral sandbox; actions beyond drafting require named-person approval.
Retention is set by you; engagement files purge at close. At exit: full export in open formats, then certified deletion within thirty days, confirmed in writing.
Access, answers and administrative actions are logged, tenant-scoped and exportable to your SIEM. Xolarc-side access to a tenant is name-listed, time-bounded and recorded.
Mandatory code review with dependency and static scanning on every change. Independent penetration testing scoped to the tenant boundary; a published vulnerability disclosure channel.
A documented response plan with notification clocks: six hours’ incident notice and seventy-two hours for the detailed report, with contractual commitments to match. Encrypted, restore-tested backups.
Data processing agreement with Standard Contractual Clauses; DPDP-first, GDPR-ready. A published, versioned sub-processor list with thirty days’ notice of change.
Security review documentation, shared under NDA.
Containerised from day one — the same platform, wherever your obligations put it.
A dedicated tenant run by us, in the region you choose. Nothing to stand up, nothing to maintain — live in days.
Deployed into your own subscription — your VPC, your region, your controls, your keys. Your security team keeps it on their side of the line.
Containerised for your own data centre — air-gappable, for regulated work and closed networks where everything must stay in the room.
The evidence pack is shared under NDA, with a direct security contact.