Manual compliance review doesnt scale
Todays marketing reviews are:
- Manual. PDFs, decks, and webpages are checked line-by-line by humans.
- Inconsistent. Different reviewers flag different issues on different days.
- Hard to audit. Months later, its unclear who approved what, against which rules, and why.
For Canadian registrants, this shows up as:
- Slow turnarounds for marketing and performance materials
- Subjective judgments with no structured trail
- Stress during exams: can you prove this was reviewed against NI 31-103 / NI 81-102?
A compliance review engine, not just an AI chatbot
RegSense ingests your marketing materials and runs them through a deterministic review pipeline:
- Document analysis. Detects performance claims, yield language, benchmark comparisons, and required disclosures.
- Rule mapping. Maps each finding to NI 31-103, NI 81-102, and relevant CSA guidance.
- Structured findings. Produces a machine-readable record of what was flagged, why, and how it was resolved.
Outcome
- Faster reviews
- Less back-and-forth with marketing
- A defensible record you can show to regulators and internal audit
What a RegSense review looks like
From flagged issues to audit-ready output.
Example input
A fund facts sheet with:
- 1-, 3-, and 5-year performance numbers
- Top quartile performance language
- A benchmark comparison vs. S&P/TSX
RegSense findings (simplified)
-
Finding #1 Performance claims
Detected 3 references to historical returns. Mapped to NI 81-102 sections on performance presentation. Status: requires confirmation of standard calculation and disclosure wording.
-
Finding #2 Top quartile claim
Detected comparative performance language. Mapped to CSA guidance on misleading performance claims. Status: requires supporting data and clarifying disclosure.
-
Finding #3 Benchmark comparison
Detected benchmark name and index comparison. Mapped to NI 31-103 guidance on fair presentation. Status: OK, pending reviewer confirmation.
All findings are stored as structured data and linked to the document, reviewer, and timestampnot buried in email threads.
Built as a verifiable compliance system
RegSense is designed for regulated environments, not demo-ware:
- Document UI integration. Works with tools your teams already use (e.g. Google Docs / Slides via Apps Script or webhooks).
- Governance layer. Every review event is written to an event ledger for traceability.
- Deterministic outputs. The same document and rules produce the same findings, every time.
- Evidence storage. Findings, decisions, and artifacts are stored in a structured way for future exams.
Not just AI for compliancean opinionated system for evidence you can stand behind.
Built for Canadian registrants
RegSense is designed for:
- Portfolio Managers (PMs)
- Investment Fund Managers (IFMs)
- Exempt Market Dealers (EMDs)
Use cases:
- Pre-clearance of marketing and performance materials
- Periodic review of websites, decks, and factsheets
- Preparing for exams and internal audit with a clean evidence trail
From ad-hoc review to structured, reproducible compliance
Before RegSense
- Manual reviews in email and PDFs
- Reviewer-by-reviewer judgment
- Approvals scattered across drives and inboxes
- Stress during exams: can we reconstruct what happened?
After RegSense
- Automated detection of key issues and references
- Findings mapped to specific rules and guidance
- Centralized log of reviews, decisions, and evidence
- Clear story for regulators: what was reviewed, when, and against which rules
Join the early access cohort
Were preparing pilot deployments with a small set of Canadian PMs, IFMs, and EMDs.
If you have recurring marketing and performance review workloads and care about deterministic, audit-ready evidence, wed like to talk.
Request Early AccessTell us about your firm and review process. Well respond with fit, timelines, and a proposed pilot scope.