Trust, security and compliance status

This is the page CTOs spend the most time on, so it is written as a board-meeting handout rather than a marketing page. Every status carries a date. Engineering evidence, operational configuration and external acceptance are separate.

Last reviewed 2026-10-15

Engineering and external review status, 15 October 2026 Bangkok; pre-revenue, no customers
Review or pledgeStatusEvidence boundary
SOC 2 Type IINo report evidencedNo completed independent audit is represented here.
FERPA, COPPA and state student-data lawsQualified review requiredTenant access, consent and export controls support a review. Architecture and passing tests do not establish legal compliance; school circumstances and agreements must be reviewed.
Student Privacy Pledge 2.0Owner verification pendingEarlier copy described a signed pledge. Confirm the signed record and applicable scope with Ross before relying on that status.
1EdTech privacy, security and accessibility rubricsInternal preparationNo award or published official assessment is claimed. Policy vetting, supplier self-assessment and certification are different processes.
Other planned certifications and filingsOwner verification pendingNo Texas filing, iKeepSafe award or NIST conformance assessment is established by this engineering review.

Start here: we are independently operated, and what that costs you

Kastr has no district in production. There is no SOC 2 report, no uptime history, no reference customer and no customer production history to substantiate service reliability. If your evaluation requires any of those, we fail your review today, and the honest thing is to tell you on the first screen rather than on the fourth call.

What we can give you instead is everything that does not require a customer to exist: the architecture, the code-level security controls, reviewable data-processing drafts, implemented-flow inventory and dated statements about what has and has not been audited. That is a smaller set of evidence than an incumbent can offer. It is also verifiable, which a logo wall is not.

The three questions a security reviewer usually asks first, answered up front. Do you have a completed SOC 2 Type II report? No completed report is evidenced; actual audit timing requires owner confirmation. Do you support school SSO or MFA? School accounts use magic links; school SSO/MFA is not implemented. The separate operator console uses password plus authenticator codes. Can we see an uptime chart? No, there is no production uptime to chart, and there is no status subdomain, because publishing a green dashboard for a system nobody uses would be theatre.

The security model, control by control

These are the controls that exist in the code today, described precisely enough that your team can test each one during a review.

Tenant isolation

Every district's data sits behind Postgres row-level security, and the application connects under a non-owner, DML-only database role set per transaction. That detail is the whole point: table owners bypass RLS in Postgres by default, so a design where the app connects as the owner has row-level security that can be silently defeated by any query the ORM decides to run. The tenant transaction switches to the non-owner role; privileged database and backup access remain separate operational boundaries.

Isolation is hierarchy-aware, a district sees its own rows and its child schools' rows, and never another district's. It fails closed: with no current organisation set on the session, every policy evaluates against NULL and returns zero rows, so the failure mode of a bug is an empty screen rather than somebody else's students. A cross-tenant leakage suite runs in CI against real Postgres 16 on every commit, including a regression check specifically for the owner-bypass case.

The audit log

Per-organisation, tamper-evident, hash-chained: each entry stores sha256(previous_hash ‖ canonical(event)), and the chain is serialised under a transaction-scoped advisory lock so it stays correct under concurrent writes. Append-only is enforced at two independent layers, REVOKE UPDATE, DELETE at the grant level, and RLS policies that permit only INSERT and SELECT with no UPDATE or DELETE policy defined at all. The chain detects accidental or partial tampering; an owner-level actor could recompute it. Independent anchoring or a secret outside the database would be needed for stronger non-repudiation.

Authentication

School accounts use magic links. The token is 32 random bytes with a 15-minute TTL, single-use, consumed atomically so that a mail-client prefetch cannot burn it and a replay cannot reuse it. It is bound to the device that requested it, by comparing a sha256(token) cookie with timingSafeEqual. There is no account enumeration, a token is issued whether or not the address exists, so the response cannot be used to test which staff emails are real. Session metadata includes an IP-hash field; host access logs and external recipients require separate review.

For school accounts there is no SAML, OIDC, Google or ClassLink sign-in, MFA or password. The separate operator console uses passwords plus authenticator codes and short revocable sessions. For a district standardised on Entra ID or Google Workspace, that is a real operational difference and a legitimate reason to choose somebody else this year.

The API surface

Organisation-scoped REST endpoints use RLS and scoped authorization, with API keys stored as SHA-256 hashes and the plaintext shown exactly once at creation. Rate limiting runs at four tiers, per key, per SCIM key, per organisation, and per IP for unauthenticated traffic, with correct X-RateLimit-* and Retry-After headers. Atomic Postgres counters coordinate these limits across application instances and retain active windows through restarts. Sign-in requests also have shared limits per address and per IP.

Outbound webhooks are signed Stripe-style, with an HMAC-SHA256 over "${timestamp}.${rawBody}", a DNS-resolving SSRF guard that rejects loopback, RFC1918 and link-local destinations after resolution rather than by string matching, exponential backoff, and automatic endpoint disable after ten consecutive failures. Roster connector configurations hard-reject literal credentials pasted into them , anything matching a known key prefix is refused, and you must use a reference pointer instead. The dedicated OneRoster REST connection is separate: its client secret is accepted in the connection form and stored encrypted with AES-256-GCM.

Encryption and transport

Application secrets use authenticated encryption and production integration requests require HTTPS. Actual database, object, backup and host encryption settings and endpoint TLS versions require deployed evidence; this page does not certify every storage or transport layer.

Data ownership, export and retention

Confirm ownership, permitted use, exit terms and any fees in the actual school agreement. Earlier references to specific MSA/DPA clauses require owner verification; a marketing page does not establish a signed contract.

An operator-run scoped export and exit workflow is implemented and synthetic-tested. It provides portable records, a manifest, checksums and authenticated expiry-bound downloads; attachment bytes must resolve or export fails. Scope exclusions are explicit, including shared district records and credentials. This is not a promise that every record is available through a one-click self-service export.

Pause and exit revoke future work and relevant credentials. Already submitted provider traffic cannot be assumed recalled. Export access expiry is not artifact erasure; exit completion does not by itself delete primary storage or backups. Irreversible deletion requires separately reviewed authority, retention and legal-hold decisions.

Record classes requiring an agreed retention schedule
Record classCurrent implementationDecision needed
Roster, student and shared family recordsWithdrawal and retirement preserve recordsSchool obligations, shared ownership, live deletion and backup expiry
Messages, forms and absence detailsNo general purge schedule evidencedPurpose, legal holds and school-approved retention
Delivery, consent and audit historyRetained; audit is append-only for runtimeRetention and lawful exceptions by record class
Unapplied roster previewsExpired previews have bounded cleanupVerify deployed scheduler and actual cleanup
Export artifactsDownload expiry blocks accessStored-artifact purge and copies held by recipient

Earlier multi-year figures were design proposals, not a legal retention determination or an enforced general purge schedule. Agree retention, deletion and backup expiry for your school before onboarding. Service pause and account closure do not establish that all records have been erased.

Sub-processors

The implemented processing paths and browser dependencies reviewed in this source inventory are listed here. This source inventory has been reconciled with implemented flows. Actual account/configuration and agreement evidence remains to be reviewed; static product settings summaries are not an approved vendor register.

Implemented processing paths and dependencies reviewed 15 October 2026 Bangkok; actual accounts and terms require confirmation
ServicePurposeData potentially processedAccount / region status
Supabase / school PostgreSQLSchool data storageStored application and roster recordsActual region, backup and contract evidence pending
Netlify / application hosting; Railway / console databaseApplication hosting and separate operations recordsRequests/application processing; customer contacts and health metadataActual enabled services, access and terms pending
ResendAccount and school emailAddresses, content and short-lived account linksActual account, tracking settings and terms pending
Twilio; Bandwidth adapter also implementedTenant-routed SMSPhone numbers, message content, delivery and command metadataTwilio is the working selection; each sender/account needs readiness and approval
DeepL free or paid APITranslationSubmitted text, which may contain personal informationPlan, retention, training-use terms and processing location pending
Google FontsBrowser fonts on app and websiteNetwork/request metadataRegion and data-handling terms unverified
Marketing analytics operatorWebsite analytics scriptBrowser/request metadata; remote script behavior unverifiedAccount, identifiers, retention and downstream use unverified

This is an inventory of supported paths and browser dependencies, not confirmation that every vendor is active for a school. Review the approved accounts, recipients, processing locations, retention and agreements before onboarding. A configured SIS, webhook, alert or backup recipient needs its own approved entry. No universal no-training, no-tracking or all-US processing guarantee is inferred.

Incident response, and what we will not chart for you

Earlier copy proposed the following response targets, which require reviewed staffing, school terms and operational evidence: Severity 1 means the service is down, data has been lost, or there has been a security breach; we target acknowledgement within 15 minutes. Severity 2, major degradation, slow delivery, partial outage, targets one hour. Severity 3, a minor or isolated issue, targets one business day. Incident notice triggers, deadlines and recipients must be agreed for applicable school obligations; a generic 72-hour figure is not every legal or contractual requirement.

These are targets from a team with no production traffic. Treat them as proposed agreement terms, not a verified response record or evidence of continuous staffed cover.

On uptime figures, and how to test any vendor's. We publish no availability percentage, because a percentage over zero months of production is a made-up number. A public status page goes up at our first district cutover, and its history will start empty and stay honest. If a competitor shows you 99.99%, the useful follow-up is not "is that true", it is "over what window, measured by what probe, and may we see the incident log behind it, including the ones you classified as degradation rather than outage". Availability figures are almost always technically true and rarely answer the question you meant.

Ask Ross for the verified security reporting route. Do not include student records or secrets in a report. Confirm monitored intake and the agreed response expectation during onboarding. No paid bounty programme is evidenced; earlier advertised bounty amounts were aspirational rather than funded and have been removed.

Procurement review documents

Ask Ross for the current reviewable agreement, data-processing terms, approved subprocessor/account register and technical evidence. Earlier document version/page-count and same-day delivery claims require owner confirmation; this source review did not inspect executed MSA/DPA documents.

Legal review, production backup/restore, live provider tests, human accessibility/language acceptance and school sign-off remain separate gates. A passing synthetic build does not close them.

Questions people actually ask

Is Kastr independently audited or certified?

No completed independent security audit, 1EdTech award or production-school acceptance is represented here. Internal engineering tests and review preparation are separate evidence.

Where does district data live, and who can see it?

School records use PostgreSQL with tenant-scoped non-owner transactions and school/family permission checks. Hosting regions, storage encryption, backup access and approved recipients must be verified against the actual school configuration.

Does Kastr support school SSO or MFA?

School accounts use browser-bound short-lived magic links with revocable sessions. School SSO/MFA is not implemented. The separate operator console uses passwords plus authenticator codes; this does not make school login MFA.

Can we export our records?

An operator-run scoped portable export is implemented and synthetic-tested, with manifest/checksums and expiry-bound authorized download. Attachments and scope exclusions are explicit. Confirm school agreement terms and request handling; no universal self-service or same-day guarantee is made.

How long does Kastr keep student data?

Agree record-specific retention, legal holds, live-store deletion and backup expiry before onboarding. General purge enforcement is not established. Withdrawal, login deactivation, service pause and export access expiry do not mean all records have been erased.

Published pricing. Current capabilities included.

$3.50 per student per year under 5,000 students. No tiers or add-on modules. Normal messaging is included under a published fair-use allowance, with transparent cost recovery only above it.