Operations Console

Assess case risk, review queue pressure, and explain the Salesforce architecture.

Static demo online
GitHub Pages mode This static app runs entirely in the browser. In the full architecture, Salesforce LWC and Apex call the FastAPI scoring service and persist results in Salesforce records.
Open Cases 142 12 assigned today
High Risk 18 3 require manager review
Overdue 11 Oldest open case: 9 days
Avg Resolution 18h Rolling 30-day sample

Case Queue

Sample worklist for analyst triage.

CASE-1042 | Payment Fraud

Bryan | High-value transfer after failed logins

HIGH
CASE-1041 | Login Issue

Sarah | Password reset and account recovery

LOW
CASE-1040 | Data Access

Alex | Unusual access from monitored country

MEDIUM

Risk Distribution

Current demo queue by assigned risk level.

High
18
Medium
31
Low
93
New case received Customer reported an unauthorized transaction.
Risk assessment calculated Current score: 100, level: HIGH.
Recommended action Escalate to fraud investigation and create a remediation action.

Operational Signals

What an analyst should look at before opening a case.

Queue priority High-risk and overdue cases should be reviewed before low-risk support cases.
Risk evidence Failed logins, country risk, and incident history explain why a score changed.
Human approval The app recommends a next action, but the analyst still decides what to do.

Risk Assessment

Enter case attributes and score the browser rules engine.

Assessment Result

Decision support output from the rules engine.

HIGH
100
CASE-1042 | Payment Fraud

Escalate to fraud investigation and create a remediation action

High transaction valueTransaction amount exceeds 10000.
New accountAccount age is under 90 days.
Multiple failed login attemptsFailed logins are 5 or higher.
High country riskCountry risk is marked HIGH.
Previous security or fraud incidentPrior incidents are present.

How This Works With Salesforce

The static page demonstrates the workflow; Salesforce is the intended system of record.

1
Analyst opens a Case in Salesforce The real implementation uses Case, Risk_Assessment__c, Evidence__c, and Remediation_Action__c metadata from force-app.
2
LWC sends scoring inputs to Apex riskAssessmentForm calls RiskAssessmentController, which delegates to service classes for scoring and persistence.
3
Apex calls FastAPI through a Named Credential RiskScoringClient posts JSON to /api/v1/risk/score in the backend version.
4
Salesforce stores the result Apex creates Risk_Assessment__c, updates Case risk fields, and Flow can move high-risk cases into remediation.

Local-To-Salesforce Mapping

Use this when explaining the project in a resume or interview.

Static UI Dashboard, queue, score form, and user manual. Salesforce equivalent: Lightning Web Components on app/home/record pages.
FastAPI POST /api/v1/risk/score calculates deterministic risk in backend mode. Salesforce equivalent: external service called from Apex via Named Credential.
Rules Engine Transaction, account age, login, geography, and incident rules. Salesforce equivalent: result saved as Risk_Assessment__c with scoring factors.
Workflow Timeline and recommended action shown to analyst. Salesforce equivalent: Flow escalation and remediation action tracking.

User Manual

Quick operating guide for the GitHub Pages demo.

Workflow

  1. Open the dashboard and review queue pressure.
  2. Choose Risk Scoring from the left navigation.
  3. Enter the case attributes or load a sample case.
  4. Click Score Case to run the browser scoring rules.
  5. Review the score, level, factors, and recommended action.

Scoring Fields

  • Transaction Amount above 10000 adds transaction risk.
  • Account Age below 90 days adds account maturity risk.
  • Failed Logins of 5 or more adds security risk.
  • Country Risk HIGH adds geographic risk.
  • Previous Incidents above 0 adds history risk.

Reading Results

Scores from 0 to 29 are LOW, 30 to 59 are MEDIUM, and 60 or above is HIGH. The recommendation is decision support for an analyst; it does not automatically close, approve, or escalate a real Salesforce case.

Salesforce Mode

In a deployable Salesforce org, analysts would use LWC screens instead of this static HTML page. Apex would receive the same inputs, call the FastAPI service through RiskScoringClient, write Risk_Assessment__c, and update the related Case risk fields.

Current Limits

GitHub Pages cannot run Python or Apex. This static app demonstrates the workflow and scoring behavior. The full Salesforce version still requires an org that supports Apex deployment and has custom-object capacity.