AI × CRM · Founder project•2025 – Present
Anveal
Anveal is pre-send governance for regulated CRM teams. Teams upload redacted campaign evidence, and Anveal checks audience logic, suppression, counts, timing and personalisation, then returns a defensible governance report with findings and fixes.
Visit Anveal
Why I built it
Every costly CRM mistake I've seen shipped through a green-lit checklist.
After seven years running CRM in fintech and insurtech, I knew the pattern: pre-send QA is hours of manual eyeballing, it depends on who checks and how tired they are, and it breaks exactly when volume grows. AI is multiplying campaign volume right now, so the checking side has to become systematic too. Anveal reviews the same evidence the same way every time, and writes down what it found.
How it works
Deterministic checks where certainty exists, reasoning models where it doesn't, and an audit trail for both.
Intake is redaction-first, so no customer data needs to leave the team. Counts, suppression and timing run as deterministic checks; audience logic and personalisation risks go through reasoning models. The output is a structured report: verdict, material findings by severity, recommended fixes and a record you can share and keep. I designed, built and shipped it solo, end to end.
Outcome
A live governance workflow that turns redacted campaign evidence into a review record teams can act on and retain.
I took Anveal from product framing through interaction design, application architecture and production delivery. The result is a working end-to-end review path: structured intake, deterministic and model-assisted checks, material findings, recommended fixes and a shareable report.