Security & data sovereignty summary.

A 2-page brief for IT, security, and procurement reviewers. Covers our certification posture, jurisdiction, AI routing controls, and the contractual commitments that travel with every customer engagement. The full Statement of Applicability and audit reports are available under a short mutual NDA — contact aj@wingrants.ai.

§1 Provider & legal

§2 Certification & audit

§3 Data residency & hosting

• European Union-hosted application by default. Multi-tenant deployment runs on AWS Frankfurt (eu-central-1) with the application hosted in the European Union. Model inference routing depends on the mode or deployment you choose.
• Enterprise deployment options. Enterprise Lite keeps the application and data on-premise, with approved sovereign or customer-approved inference. Enterprise Campus Compute ships the full platform as a self-contained Docker Compose bundle with open-weight model weights included for fully local or air-gapped inference, requiring no runtime third-party model calls. Customer chooses where it runs — on-prem, sovereign EU cloud, or fully offline.
• Sub-processors. Minimised list, EU-resident where possible, published in the DPA. Annual reassessment under ISO A.5.19–5.23.
• Backups. Immutable WORM backups, 90-day retention, geographically distributed across EU. RPO 1h · RTO 4h. Quarterly DR tabletop exercises.

§4 AI routing & model controls

• Configurable model routing. Bring-your-own-keys (BYOK) can use commercial model routing through OpenRouter, including United States model providers when enabled. Managed routing limits routing to open-source, European, or customer-approved model options where available — safer and more controlled for sensitive proposals. Campus Compute Mode runs inference locally on customer infrastructure for Enterprise Campus Compute deployments. Self-Service supports bring-your-own-keys (BYOK); AI Enablement adds a tenant-level routing policy.
• Open-weight options. Where open-weight routing applies, the platform routes between Llama 4 Maverick, Qwen 3.6, Mistral Large 3, DeepSeek V4, Gemma 4, and gpt-oss-120B depending on task. Enterprise Campus Compute runs approved open-weight models with no runtime third-party model calls.
• Version pinning. Models pinned to specific versions; upgrades require written customer approval on Enterprise. The old model stays available in your tenancy until you choose to retire it.
• Inference routing by plan and mode. Model inference routing depends on the mode or deployment you choose. Restricted and Enterprise routes keep inference within approved European or customer-approved endpoints; Campus Compute keeps inference fully local. Regional pinning available on Enterprise.
• Configurable model routing. Bring Your Own Key (BYOK) is supported when customers want to route usage through their own approved model accounts or endpoints on AI Enablement and Enterprise. Hold-your-own-keys (HYOK) reserved for air-gapped Enterprise Campus Compute deployments.

§5 Identity, access, encryption

§6 GDPR & regulatory alignment

• Role. Data Processor on behalf of the Customer. Standard EU DPA available; supports Customer-supplied DPA on request.
• Lawful bases. Customer determines lawful basis for processing in their own role as Controller. Sub-processor list and international transfer schedule published in the DPA.
• Data subject rights. Self-serve export in JSON and DOCX; admin-initiated full export and delete on request. Deletion timelines (including backups) documented in the DPA.
• Incident response. Playbook-driven; 4h customer notification SLA; 72h regulator notification per GDPR Art. 33.
• EU AI Act. WinGrants AI is classified as a limited-risk AI system under the EU AI Act. Output labelling, transparency obligations, and human-oversight requirements are designed in.
• NIS2. Controls aligned with the NIS2 Directive risk-management framework via the ISO 27001 ISMS.

§7 Contractual commitments

• Enterprise Data-Flow Guarantee — your data-flow map is defined before launch. Enterprise Lite keeps application data in your environment and routes inference only to approved endpoints. Campus Compute keeps both application data and model inference inside your perimeter. Written into the DPA, audited annually by your IT, verifiable in our open codebase.
• Campus Compute Open-Weight Guarantee — for Campus Compute deployments, model inference runs on approved open-weight models. Enterprise Lite may use approved inference endpoints. We will never swap in a model dependency outside the agreed route without your written approval.
• Exit Guarantee (Enterprise) — on termination you keep the Docker bundle, your database export, every custom prompt, and, for Campus Compute, the open-weight model weights. Stand up an offline instance forever.
• AI Enablement Configuration Guarantee — your prompts, scorers, templates, workflow configuration, and isolated database exports remain yours under contract.
• 100% Satisfaction Guarantee (Self-service only) — a one-time full refund under the Fair-Use Policy in the Terms; an approved refund permanently closes the account. Licence plans (AI Enablement, Enterprise) carry a Submission-Ready Guarantee instead: we iterate until your team is producing submission-ready drafts, under the fair-use terms of the Order Form. Licence fees are non-refundable.