Entry product · Process map
Smart Diagnose — map FINTRAC ops against real cases
Best when you need a structured gap and runway view before Blueprint or Pilot. Profile stack and pain, complete the advisor process map, and leave with scored exception workflows. For a live desk walkthrough use Case Desk; for a short fit check use Free Assessment; for a customer-specific proof request Workflow Proof.
Settlement boundary
TrustField does not itself provide custody or settlement services. Applicable custody and settlement activities remain with the customer’s designated providers.
Expert session · buildable workflow map
Smart Diagnose · adaptive map
Synthetic public demo
Rules review demo
Not a private customer ticket. Not a verified Brain Circle claim.
Loading demo…
Rules review (not a verified Brain Circle)
After you save a map, TrustField runs a deterministic six-role rules review (qualifier, auditor, critique, advisor, machines, workflows) in parallel, then synthesizes while preserving disagreements that match each role's stored verdict. Review-mode badges show Rules review, LLM review not configured / completed, and Human review pending. While LLM is disabled, citations come from an official-source rules/citation pack (not RAG). External LLM stays off until Privacy, Data Processing, and the counsel DPA allow it — rules never let an LLM decide FILE/NO-FILE, invent obligations, auto-submit, override human gates, or silently change scores.
Why this exists
Automation design fails on shallow quizzes. This session walks Compliance from identification through submit and archive, flags report-type drift / GRC shelfware / missing exception SOPs, and scores you against composite cases — including MSP multi-client desks that still run on spreadsheets.
Advisor discovery + buildability
From identification for review/reporting → register, send, and final archive. Includes triggers, report types, systems, manual steps, copy/paste, controls, approver, FINTRAC send path (FWR / API / paper modeled separately), archive, exception/amend, bottleneck, and redacted artifacts. Buildability questions make the first workflow constructible and testable. Exception cases matter as much as happy-path cases. Recommendation is answer-driven after applicability gates — missing/conflicting data is a common hypothesis, not a preselected pilot.
Real-case patterns in the engine
- MSB · EFTR alert flood → Excel re-key → portal
- Exchange · wallet / chain alert → STR narrative
- Remittance · near-threshold cash → LCTR / aggregation
- Corrections-heavy shop · reopen tickets, no amend SOP
- GRC licensed · Excel/paper still the truth
- MSP · multi-client FINTRAC desks on spreadsheets
Built for
- MSBs and money transmitters
- Crypto / digital asset exchanges
- FX, OTC, remittance, and payout operators
- MSPs / compliance consultancies serving MSB clients (channel)
- Fintechs carrying MSB-style obligations
What you leave with
- Map quality score + cross-check flags
- Matched real-case pattern (as-is, failure modes, to-be)
- Full FINTRAC process map (identify → archive)
- Buildability lane (volume, time, roles, access, exceptions, audit, retention, success)
- Normal case + exception case samples (both required for real v1)
- All applicable ICP exception workflows assessed (not a fixed three)
- One recommended first pilot generated from your answers (or none until evidence supports it)
- Rules review with six role contracts + preserved disagreements (not a verified Brain Circle claim)
- Automation candidates + ranked production runways
- Exportable JSON brief for Blueprint / Pilot scoping
TrustField builds program operations and evidence software. It records program state and prepares payload-free handoff records; licensed partners execute movement. Humans remain the FINTRAC filers. MSP channel: partners distribute — TrustField is not an MSP operator.