NorraCRM

Norra CRM

Platform

Explore every module of Norra CRM: control tower, trader 360, challenges, risk engine, fraud graph, payout control center, double-entry finance, AI agents and governance.

Operate

Operate

Executive Control Tower

Open one screen. Know the state of the firm.

KPI groups for revenue, payouts, traders, customers, risk, fraud, finance and operations — every number carries a link to its source. The tower tells you what changed since your last visit and ranks what needs attention right now.

  • What-changed deltas vs your last visit, 24h and 7d
  • Ranked needs-attention feed across alerts, payouts, findings and SLAs
  • Live Trust Score with drill-down
  • Hourly snapshots — every comparison is reproducible
Executive control tower

Operate

Trader CRM & Customer 360

Every trader, one encrypted profile.

Identity, KYC state, devices and IPs, wallets, payments, challenges, accounts, tickets, fraud band and a full activity timeline — on one page, masked by default, permissioned to the field level.

  • Field-level AES-256-GCM encryption; unmasking is audited
  • Status workflow (restrict / suspend / ban) gated by approvals
  • Device, IP and wallet link graph per customer
  • Segments, VIP tiers, tags; GDPR export and crypto-shred deletion
Customer 360 profile

Operate

Challenges & Evaluation

Design the product. The lifecycle runs itself.

Build challenge products with phases, profit targets, static or trailing drawdown, minimum trading days and payout splits. Purchases activate challenges, evaluations track automatically against versioned rules, and funding happens only after a human approves.

  • 1–3 phase products with per-phase targets and drawdown types
  • Product versioning — every challenge keeps the rules it was sold under
  • Automatic activation on payment, evaluation on live metrics
  • Funding requires explicit approval; every funding event is audited
Challenge product configuration

Operate

Support Desk with AI Drafts

Tier-1 support, drafted by AI, sent by humans.

Inbound tickets are classified (category, sentiment, priority) against a masked context card — payout status, KYC state, account health, fraud band only. The support agent drafts a reply that stays invisible to the customer until a human approves it, or an approved auto-send policy covers the category.

  • Automatic classification with confidence and sentiment
  • Context card without PII exposure — status, never documents
  • Drafts are structurally unsendable without approval or an approved policy
  • SLA tracking with breach alerts; macros; injection-flagged tickets never auto-send
Support ticket with AI draft awaiting approval

Control

Control

Risk Engine & Surveillance

See risk the moment it forms.

Account and company risk scores recomputed continuously from drawdown proximity, leverage, concentration, behaviour flags and volatility. Exposure maps by symbol, country and cluster. Surveillance detectors classify trader behaviour from martingale to copy trading.

  • Account + company scores with component breakdowns
  • Exposure by symbol / country / linked-account cluster with limits and approvals
  • Detectors: martingale, grid, revenge, lot spikes, news & weekend trading
  • Trade-similarity engine surfaces copy trading across accounts
Risk exposure dashboard

Control

Visual Rule Engine

Your rulebook, as executable policy.

Risk and trading rules are data, not code: drafted against a typed fact catalogue, simulated against history before publishing, published under four-eyes approval, and immutable once live. Every evaluation is recorded and replayable.

  • Draft → mandatory simulation → two-approver publish
  • Immutable published versions; changes create new versions
  • Fail-closed evaluation — a missing fact never passes a rule
  • Full evaluation records: replay any decision input-for-input
Rule engine with simulation

Control

Fraud Intelligence

The farm can’t hide from the graph.

Signals from devices, IPs, wallets, KYC identities, payments, behaviour and referrals feed an explained fraud score with recommendation bands. Connected components become clusters; a relationship graph lets analysts walk every link. Scoring never enforces — enforcement is an approval.

  • Explained scores: every point traces to named signals
  • Cluster detection across devices, IPs, wallets and identities
  • Interactive relationship graph with similarity edges
  • Recommendation bands from allow to escalate — never auto-ban
Fraud relationship graph

Control

Investigations & Case Management

From suspicion to decision, on the record.

Cases with typed evidence lockers, SLA tracking and agent assistance. Enforcement — restriction, suspension, freezing, payout holds — executes only through approvals, and every step lands in the case timeline.

  • Evidence locker: similarity, links, signals, findings, notes, files
  • SLA clocks with breach alerts
  • Approval-gated enforcement; AI agents can open cases, never close accounts
  • Closed cases become searchable organisational memory for the AI layer
Investigation case with evidence

Control

KYC & Compliance

Compliance that keeps up with your funnel.

Provider-agnostic KYC flow with levels, webhook verification, duplicate-identity fingerprints, sanctions and PEP screening, and approval-gated overrides. Wallet screening gates every payout.

  • KYC levels with automatic approval routing and manual review queues
  • Duplicate identity detection across document fingerprints
  • Sanctions / PEP screening with hold automation
  • Overrides require approval; every decision is audited
KYC review queue

Money

Money

Payments & Crypto Invoicing

Crypto-first checkout that reconciles itself.

Per-invoice HD deposit addresses on five chains with exact, under-, over- and duplicate-payment handling; wallet screening on inbound funds; a fiat PSP rail with verified webhooks; coupons, refunds through approval, and chargeback management with prediction.

  • USDT/USDC on TRON, Ethereum, Polygon, Solana, BSC
  • Underpayment, overpayment and duplicate deposits handled explicitly
  • Refunds flow through approvals and post contra-revenue automatically
  • Chargebacks post losses, close challenges, hold payouts — and predict themselves
Payments and crypto invoices

Money

Payout Control Center

Every payout scored, explained and approved.

Each request is scored by a deterministic model across 27 factors — fraud, wallet, identity, trading behaviour, history — into an explained decision: auto-approve within policy caps, single review, enhanced review with two approvers, or hold. The sender must differ from the requester and every approver, and funds settle only after on-chain confirmation, posting into the ledger.

  • Explained scores: every factor, weight and blocker visible
  • Auto-approval only inside policy gates — never a first payout
  • Four-eyes on high value; separation of duties on send, enforced by the database
  • Chain-confirmed settlement; liability and cash posted double-entry
  • Any decision reconstructable months later — facts, rules, votes, transaction
Payout control center

Money

Finance, Ledger & Economics

Books that balance by construction.

An immutable double-entry ledger underlies every money movement. P&L, daily reconciliation against on-chain reality, cost tracking, forecasts, cash-requirement planning — and a challenge-economics simulator that replays your actual trader history under changed rules before you change them.

  • Double-entry ledger with balanced-transaction enforcement in the database
  • Daily reconciliation: ledger vs chain vs PSP, with persisted runs
  • Revenue, payout, customer and cash forecasts with confidence bands
  • Economics simulator: "profit target 8→10%" replayed against real history, with backtesting
Challenge economics simulator

Money

Affiliates & Marketing Intelligence

Growth you can audit.

Click tracking, multi-model attribution, commissions that accrue as reversible ledger entries and claw back on refunds and chargebacks, affiliate quality scores with self-referral detection, and marketing analytics down to contribution margin and risk-adjusted LTV.

  • Attribution: last-click, first-click, linear, time-decay — your choice
  • Commissions mirrored in the ledger; reversals and clawbacks automatic
  • Quality scores: fraud exposure, traffic authenticity, self-referral flags
  • Contribution margin and risk-adjusted LTV per channel, campaign and affiliate
Affiliate profile with quality score

Intelligence

Intelligence

AI Agent System

Thirteen specialists. Zero trust. Full records.

Payout review, fraud triage, compliance, support, risk, security, trading surveillance, an internal auditor, a data analyst and executive briefing agents — each with an explicit tool scope, data scope, action limits and token budget. Every run keeps its complete tool-call trace; every decision keeps its evidence.

  • Zero-trust scopes: an agent cannot call a tool it wasn’t granted — refusals are alerted
  • Forbidden actions (ban, ledger, KYC, payout send) cannot even be registered as tools
  • Approval requirements computed by policy, never by the model
  • Accuracy dashboard: agreement rates, overrides, calibration per agent
  • Hourly token budgets that fail closed
AI agent decision with evidence

Intelligence

AI Command Center & NL-BI

Ask the firm a question. Get a provable answer.

"How many payouts were rejected last month, by country?" The data analyst translates questions into guarded, read-only SQL over a governed semantic layer, shows you the query, checks every number in the answer against the result cells — and says "DATA NOT AVAILABLE" rather than guessing.

  • Guarded SQL: single read-only statements over approved views, 5-second timeout
  • Every number programmatically verified against the result set
  • Daily executive briefing and weekly board report, every figure traceable
  • Internal auditor sweeps: MFA gaps, stale keys, scope drift, control weaknesses
NL-BI answering a question with SQL shown

Assurance

Assurance

Governance, RBAC & Audit

Control who does what. Prove it later.

Sixteen roles and 90+ granular permissions with per-grant constraints (amount caps, case-linked access, masked views); MFA and step-up for critical actions; API keys with scopes; and an append-only, hash-chained audit log verified continuously.

  • Permission constraints: approval caps by amount, PII access only case-linked
  • TOTP MFA required for staff; step-up re-auth on critical actions
  • Hash-chained audit log — tampering breaks the chain, loudly
  • Approvals with database-enforced separation of duties
Trust score and governance view

Assurance

Operations Health & Data Quality

Know your platform is healthy — continuously.

Every integration probed on a 30-second cycle with SLOs and circuit breakers; webhooks retried and dead-lettered; incidents with impact analysis and post-mortems; a 24-check data-quality suite scoring the data your decisions run on; insider-risk analytics watching the watchers.

  • Integration health with circuit breakers and recovery alerts
  • Incident management: timelines, impact (customers, revenue at risk), post-mortems
  • 24 data-quality checks across completeness, consistency, freshness, uniqueness, reconciliation and validity
  • Insider signals: unusual volume, off-hours activity, excessive overrides, permission drift
  • Live Trust Score with hard penalties and a self-audit loop
System health and data quality

Your traders see every operational mistake.

Give your team the platform that makes the mistakes hard to make — and everything provable afterwards.