RiskBoard
RiskBoard
Transaction Risk cabinet: score 0–100 with clear reasons
RiskBoard imports Excel/CSV payment operations, calculates risk_score 0–100 with reason codes, and provides a review workbench, dashboard, and AI Analytics.
- Challenge
- Analysts needed a fast way to score a payment-operations file and surface high-risk cases with clear reasons — without replacing AML systems.
- Solution
- Built RiskBoard: Excel/CSV import, risk score 0–100 with reason codes, a review workbench, dashboard, and AI Analytics for a selected company and period.
- Result
- The demo cabinet lets you upload operations, tune rules, and run a review cycle at riskboard-demo.djasen.studio.
RiskBoard is a Transaction Risk cabinet for analysts and compliance teams: upload payment operations, get a risk score 0–100 with clear reasons, review cases, and analyze patterns for a selected company and period.
It evaluates operational/payment risk signals. It is not an AML / financial monitoring system and does not replace legally required compliance tools.
Platform capabilities:
• Import — Excel/CSV; required fields: tranId, date, time, amount, merchant, bankName, tranType; full PAN is rejected, only panMask is accepted; other columns are kept in extra_fields
• Scoring — risk_score 0–100 and reason codes; core rules: amount above threshold, night-time window, merchant repeat; extra/custom rules per company (risky type, panMask repeat, amount velocity, round amount, and more)
• Transactions workbench — filters by merchant/bank/type/score/tags/period, tags, review status, bulk actions, and export
• Dashboard — charts and concentration by merchants, banks, high-risk samples
• AI Analytics — risk-pattern analysis for the selected company and period (scores, merchants, banks, reasons) plus a general AI Assistant
• API Keys — machine scoring of transaction batches without interactive login, with response field scope control
• Multi-company / i18n — companies, access permissions, UK/EN/RU languages
Key scenarios:
• quickly score an operations file and surface high-risk cases
• tune thresholds and rule weights to company policy
• tag cases and move them through review
• connect external systems via the scoring API
Stack: Django 6.1, PostgreSQL, Celery, Redis, AI providers.
Demo: https://riskboard-demo.djasen.studio/