Stitch Sherlock
Powered by Stitch

Onboard business customers with a risk score you can defend.

Stitch Sherlock collects company, ownership and trading data, verifies it against South African registries, and returns a weighted risk rating your compliance team controls.

Free while you test · FICA-aligned · Data stays in South Africa

sherlock.stitch.money/dashboard/settings/risk-weights
Risk Weights
Preview and version risk-scoring configuration
Publish v5
Factors
Revenue bandrevenue_band9.020.3%
Industry categorycategory9.020.3%
Legal entitylegal_entity7.817.6%
Sales channelsales_channel5.512.4%
Score bandsValid
LowMed lowMediumMed highHigh
020406080100

How it works

Four steps from the first merchant details to a decision your auditor can follow.

01
Estimate

The first merchant details — industry category, revenue band, legal entity, sales channels — run through your weightings for a normalised 0–100 score and a band. You own the model, not us.

02
Verify

Registry, ID, bank account and sanctions checks run in parallel. Every result is stored with its source and timestamp.

03
Collect

The risk band decides what else you need: directors, UBOs, documents and declarations — pre-filled from CIPC where we can, and proportional to the risk.

04
Decide

Auto-approve low risk, route the rest to review with the full reasoning attached. Re-run on any trigger.

RISK SCORING

Every score is a sum you can show a regulator.

No black box and no model drift you didn't ask for. Each factor carries a weight you set, each answer maps to a value, and the assessment records exactly which weights were live at the time.

Versioned rulesets — publish, compare, roll back
Simulate against real merchants before you publish
Existing assessments are never silently rescored
Ruleset v4 · live6 factors
Revenue band9.0
Industry category9.0
Legal entity7.8
Sales channel5.5
Product mix4.9
Geography4.2
LowMed lowMediumMed highHigh
020406080100

Built for South African compliance

Accountable institutions carry the obligation. We give you the evidence trail to meet it.

FICA & the FIC Act

Customer due diligence, enhanced due diligence triggers and ongoing monitoring, mapped to your RMCP.

POPIA by default

Purpose-bound collection, consent captured at source, retention schedules and subject access requests handled in-product.

Data residency

Customer data is processed and stored in af-south-1. Full export on request, deletion on schedule.

Set up your risk model in an afternoon.

Start with our template, tune the weights, publish when you're happy.