Security & Trust

SeaSquad is built for fleet operators who need to trust their software with sensitive operational, commercial, and compliance data. This page describes the security posture of the platform: how we isolate client data, control access, handle AI workloads, and run our infrastructure.

Deployment model

Every SeaSquad client runs on a single-tenant deployment. Your instance has its own compute, its own database, and its own object storage — there is no shared multi-tenant data plane where different clients' data lives side by side in the same tables or buckets. This is a deliberate architectural choice, not a configuration option: it means a bug or misconfiguration in one client's environment cannot expose another client's data, because there is no shared surface for that to happen across.

Single-tenancy also gives clients control over deployment specifics — including region placement — where that matters for their operational or regulatory requirements.

Data protection

Client data never commingles. Each deployment's database and object storage are scoped to that client alone, and application logic never queries across tenant boundaries because there is no cross-tenant boundary to query across in the first place.

• Encryption in transit. All traffic — between the browser, the mobile app, and our services, and between internal services themselves — runs over TLS.
• Encrypted backups. Backups are encrypted at rest and retained under a defined schedule, so data can be restored without extending the window of exposure.
• Signed access to documents. Certificates, inspection reports, manuals, and other files in object storage are never served from a publicly readable location. Access goes through short-lived, signed URLs scoped to a specific file and a specific request.

Access control

SeaSquad implements role-based access control across three tiers, so permissions match how maritime organizations are actually structured — from platform administrators, to company-level management, to fleet and vessel-level users.

• Vessel scoping. Individual users are scoped to the vessels they're authorized to see. A user assigned to specific ships cannot query or browse data for vessels outside their assignment, regardless of what role they hold.
• Session security. Authentication uses JWT-based sessions with support for revocation, so access can be cut off immediately when a session needs to end — not just when a token expires on its own.
• Audit trails. Administrative actions are logged, giving organizations a record of who changed what, and when.

AI and your data

SeaSquad's AI features — document intelligence, chat over fleet correspondence and records, and web-search-assisted answers — are built to respect the same data boundaries as the rest of the platform.

• Your data stays yours. AI features run against your organization's own data only. There is no cross-client data access for AI processing, consistent with the single-tenant deployment model above.
• Retrieval-grounded answers. Rather than relying purely on a model's internal knowledge, AI responses are grounded in retrieval over your organization's actual documents and records, which keeps answers traceable back to source material.
• No training on client data. Client data is never used to train or fine-tune underlying AI models. It is used only to serve that client's own requests.
• Data-residency options. For clients with strict data-residency or data-sovereignty requirements, self-hosted model deployment options are available, so AI inference can run entirely within an environment you control.

Operational security

• Infrastructure as code. Deployments are defined and provisioned as code, which keeps environments consistent and changes reviewable rather than made ad hoc against running systems.
• Continuous monitoring. Infrastructure and application health are monitored continuously, with alerting so issues are caught early rather than discovered by users.
• Backup and disaster recovery. Backup and recovery procedures are defined in advance and exercised, not improvised after the fact.
• Security review gates. Changes go through security review as part of the development process before they ship, rather than being audited only after the fact.

Reporting a vulnerability

We take reports from the security research community seriously. If you believe you've found a vulnerability in SeaSquad, please contact us at security@seasquad.ai. We ask that you give us a reasonable window to investigate and address any issue before public disclosure, and we will acknowledge and work with you in good faith on any legitimate report.