Built for high-stakes learning. Accountable by design.
Sukrat handles student learning data because the product needs it to personalise teaching. Not for advertising, resale, or model training.
What we promise every family and school.
No data sales
Student data is used to personalise learning and show progress. It is not sold to advertisers or data brokers.
No AI training on student data
Student data is not used to train third-party AI models. This is enforced in architecture and in vendor agreements.
No personal data in AI calls
Personally identifiable information is stripped before any external model call. Providers receive curriculum context and anonymised session content only.
Content built to exam standards
Practice and feedback are grounded in retrieved exam materials, not pulled from random internet searches. Off-topic prompts are blocked.
Account-level isolation
Row-level security means each student sees only their data, teachers see only their cohort, and schools see only their institution.
EU-region infrastructure
Student data is stored in the EU region. Encryption at rest and in transit across application and pipeline layers.
A system that guides, not a homework answer engine.
Sukrat teaches to the weak topic, not to the answer. Feedback traces to the specific criterion a student missed. The system scaffolds understanding through follow-up questions, enforced at the system level rather than as a prompt afterthought.
Student services are intended for users aged 13+. Guardrails screen inputs and outputs before they reach students. Generated school content is teacher-reviewable before students see it.
- Input validation and output filtering on tutoring calls
- Grounding over thousands of real exam materials
- Curriculum scope filters by course, level, and paper
- AI calls traced for quality and safety review
- Outcome claims published only when measured and permissioned
What we store, where it lives, and how long we keep it.
| Category | Approach |
|---|---|
| Learning data | Answers, mastery scores, session history, and topic maps. Used only to personalise teaching and show progress. |
| Account data | Name, email, school affiliation, and role. Used for authentication and access control. |
| Retention | Active accounts keep data for the life of the account. Deletion requests are processed within 30 days, and backups are purged on standard rotation. |
| Storage region | Primary student data in EU-region infrastructure. Encryption at rest and in transit. |
Third parties that process data on our behalf.
All vendors are bound by data-processing terms. Personal data is stripped before external AI calls.
| Provider | Purpose |
|---|---|
| Supabase | Database, authentication, row-level security |
| Vercel | Web application hosting |
| DigitalOcean | Backend API and ML pipeline |
| AI model providers | Content generation with anonymised curriculum context only. No personal data, and no training on student data. |
Full subprocessor list and data-processing terms are provided during school procurement. Request: sahmad@sukrat.ai
Full legal and policy documents.
For procurement, parent review, or implementation teams.
Privacy Policy
What we collect, why, retention, and your rights.
Terms of Service
Account use, subscriptions, and service terms.
Cookie Policy
How cookies and analytics are used on sukrat.ai.
School DPA & procurement
Data Processing Agreement, pilot structure, success criteria, and rollout support for school buyers.
Questions: sahmad@sukrat.ai for schools and DPA, or zayad.malik@sukrat.ai for technical.
Trust should be inspectable, not implied.
Explore the product, read the policies, or talk to us about school deployment.