RISK SCORING

A risk read your team can explain, not a black-box verdict

Noxtica turns browser, device, and network evidence into a score, a named risk level, a confidence value, and reasons. Your team maps them to its own actions.

Outputs your team can read together

Review the fields together. A higher score with limited confidence can call for more evidence rather than a hard block.

Noxtica

Risk level

  • Minimal
  • Low
  • Medium
  • High (example read)
  • Critical

Confidence

LimitedHighIn this example: limited.

How much evidence supported the read. It is separate from the level.

Named reasons

  • Automation and bots
  • Fingerprint tampering
  • Infrastructure abuse

Example reason families. Visible reasons follow your role and the evidence gathered.

Your policy

Noxtica supplies the read. Your policy supplies the action, set for each route. Start in observation mode.

  • Allow
  • Observe
  • Step up
  • Review
  • Block
A risk read, not a verdict: level, confidence and reasons arrive separately, then your policy actsIllustration with example labels, not customer data.
  • Score and risk level

  • Confidence

  • Named reasons

  • Console review and calibration

From evidence to an action you choose

  1. Your page

    Browser collector

    Gathers browser, device and network context. Optional collection follows your consent and admission settings.

  2. Noxtica

    Risk read

    Evidence from the page

    A score, a risk level, a confidence value and named reasons.

  3. Your server

    Server lookup

    Level, confidence, reasons

    Your backend reads the assessment with a server-held credential. The page callback and the console expose it too.

  4. Your policy

    Choose the action

    Risk bands

    Allow, observe, step up, review or block, set in your own backend. Start in observation mode.

How a risk read moves from your page to an action you choose
  1. Gather evidence

  2. Score and explain

  3. Choose the action

What a score cannot tell you

  • Evaluation access. Core scoring belongs to an activated web integration.
  • Enforcement is controlled. Challenge, limit, and block actions need controlled provisioning and explicit configuration for your tenant.
  • Estimates, not facts. A score does not prove a visitor is human, fraudulent, or safe. Output is probabilistic, and no accuracy or false-positive rate is promised. Your team makes the decision.
  • Missing evidence is not low risk. Treat unknown as unknown.

Evaluate the read against your own traffic