Dissertation Editing Center · NYC · NJ · CTOrder a review →
Dissertation Editing Center
Vol. I · No. VISecurity & privacy

Your manuscript, in five sentences.

Before the procurement-grade detail below — here's the part that matters most if you're a candidate about to upload chapter 3 tonight.

Reading room of a contemporary academic library with tall windows and rows of desks.
Reading room of a contemporary academic library with tall windows and rows of desks.Photograph · Iñaki del Olmo — Unsplash
Plain-language summary
  1. Who sees your manuscript: only the reviewing agents and you. We do not give your manuscript to a human reviewer.
  2. Where it lives: encrypted at rest in single-tenant US-region Postgres (AES-256), in transit over TLS 1.3. The manuscript bytes never leave the platform's database.
  3. Does it train any model: no. We never send your manuscript to any third-party AI provider for training. We contract with Anthropic under no-train terms; the manuscript appears only in one-shot review prompts.
  4. When it's deleted: your manuscript and review are retained for 12 months by default so you can re-download them, then purged — and you can hard-delete on demand from your dashboard at any time. Deletion is propagated within 24 hours and confirmed by email.
  5. How to delete it on demand: Dashboard → Manuscripts → ⋯ → Delete permanently. Or email support@dissertationeditingcenter.com with your confirmation ID; we respond within 1 business day.
Compliance posture
  • FERPA-aware — student-record-grade controls applied to every job.
  • GDPR posture — lawful basis: contract performance + consent. EU data subjects: write to support@dissertationeditingcenter.com.
  • SOC 2 Type II — in progress. Target audit window: Q4 2026. Trust Center launches at start.
  • Sub-processors — Anthropic (no-train), Supabase (US-region), Upstash Redis, Vercel. List maintained at /security/subprocessors.
Security & privacy

Institutional-grade controls for doctoral work.

Scholaria handles unpublished doctoral research. The platform is engineered to be safe to deploy inside a writing centre, graduate school, or research office without compromise.

Encrypted in transit & at rest

TLS 1.3 on the wire. AES-256 envelope encryption for stored manuscripts. Per-job keys, never re-used.

Retention you control

Plan-controlled retention windows. Enterprise institutions set retention at the programme level. Hard delete on request.

Single-tenant data isolation

Manuscripts are scoped to your account. No cross-customer training. No retention by the model provider.

Institutional SSO & SCIM

SAML 2.0 and OIDC for sign-in, SCIM 2.0 for provisioning. Audit logs exportable to your SIEM.

FERPA-aware controls

Built to be deployed inside a FERPA-aware programme. DPAs and BAAs available on the Enterprise tier.

Full audit trail

Every agent invocation, finding, and revision is durable, replayable, and exportable. Nothing is opaque.

Academic integrity

The first principle of the platform.

Scholaria is not a ghostwriter and is not a paraphrase laundry. It is engineered to be defensible if a chair or committee asks how it was used. Every output is auditable.

We critique. We do not author.

Scholaria never produces full replacement prose for entire sections. Findings reference verbatim excerpts and recommend changes the student decides whether to apply.

Originality assistance, not bypass

Turnitin-style overlap signals and paraphrase guidance — designed to make student work submission-ready, never to evade detection.

The QA agent has veto power

Any output that reads as AI-generated, generic, or template-driven is rejected and regenerated before it ever reaches a student.