How Maya works
under the hood —
built for compliance.

The guard runs a pattern matcher over every draft — "your rate is", "you're approved for", "you qualify", "you should choose" — and either rewrites the reply to soften the framing or escalates to a licensed mortgage advisor. Every rewrite is captured alongside the original. The rule is enforced as code, not as a paragraph in the system prompt. A jailbroken prompt can't bypass it because the rewriter sits between Maya's output and the user.

Memory isn't a single blob. Types: LONG_TERM, PREFERENCE, GOAL, LIFE_EVENT, RELATIONSHIP, EPISODIC, SENSITIVE. Sensitive memories (mental-health context, immigration status, marital trouble) are redacted in API responses, exports, and admin search — advisors see a placeholder, not the raw value. Each memory carries a confidence score, a timestamp, and a provenance link back to the originating turn. Stale or contradicted memories decay rather than accumulating noise.

Every night the harness runs deterministic checks (did Maya extract the right intent? did the compliance guard fire when it should? did the right tool get called?) alongside LLM-judged quality checks (was the empathy framing appropriate? was the math correct?). Failing evals block deploys via GitHub Actions. Scenarios are versioned in source control so we can replay any past failure and prove the fix. 100+ scenarios covering renewals, first-time buyers, stress test edge cases, multilingual flows, B-lender qualification, and more.

Every voice call is recorded with consent collected at the start of the call, stored encrypted, and tagged with a retentionUntil timestamp. A nightly sweeper hard-deletes anything past retention — not soft-delete. Every playback writes an audit row (who listened, when, from where). Every transcript share link uses a synthetic share:<prefix> identifier so client downloads live in the same trail as staff listens. Nothing happens to a recording that isn't logged. Auto-transcription with the same retention as the source audio.

All Maya data — conversations, transcripts, memory, recordings, audit logs, eval results — lives in Canadian regions. Compute runs in Canadian regions. Backups stay in Canadian regions. There is no cross-border data flow for customer content. Encryption is end-to-end: TLS 1.2+ in transit, AES-256 at rest. Keys are managed via a cloud KMS with workspace-scoped envelope encryption for the most sensitive memory categories. Models are accessed via inference-only endpoints with vendor data-retention disabled.

Your brokerage admin can mint a revocable, time-limited share link that downloads your transcript as text or JSON — no portal account required. Each link is high-entropy, scoped to one recording, and capped at 14 days by default (60-day hard maximum). Every access bumps a counter and writes to the audit log. Admins can revoke at any time. Full conversation + memory exports and deletion requests are available from your account settings — no support ticket required. Deletion runs within 30 days with conversation audit anonymization.

Every conversation, memory, recording, eval result, outreach attempt, and lead intent is workspace-scoped — enforced at the database query layer, not at the UI. Admins from brokerage A cannot see anything from brokerage B. Within a workspace, role gates control what each staff member can see: an agent sees their assigned conversations, a manager sees their team, a principal broker sees everything. Roles are configurable per workspace. SSO (Microsoft / Google) + SCIM provisioning on Enterprise. API keys are rate-limited, audited, and individually revocable.

If we discover a security or privacy incident affecting your data, we commit to notifying your workspace admin within 72 hours, providing a preliminary scope assessment, and posting a public post-mortem (no customer-identifying information) within 14 days of resolution. We run a documented incident-response playbook — versioned in source control as INCIDENT_RESPONSE.md — covering severity classification, containment, forensics, customer communication, PIPEDA regulator notification, and remediation. Security questions and responsible disclosure route to info@mortgagesquad.ca.