Onboarding

Getting started in 7 steps

A guided path through the Federal Bank AI governance workspace. Follow the steps in order for a full tour, or jump straight to the area you need.

This is a demonstration environment. Data is sample data for Federal Bank (India), and no sign-in is required — the header shows a demo user.

  1. Read the executive posture

    Start on the Executive Dashboard to see the composite AI Trust Score for Federal Bank and how compliance, responsible AI, security and audit readiness are trending.

    • Hover any chart to read exact values for the period.
    • The agent heatmap flags which dimension is dragging a score down.
    Open Executive Dashboard

  2. Browse the governed AI estate

    The Agent Registry and AI Agent Inventory list every agent, its owner, business line, model and current risk tier.

    • Filter by business line or risk tier to narrow the estate.
    • Open an agent to see its controls, evidence and recent decisions.
    Open Agent Registry

  3. Run the Demo Agent

    Pick a scenario (RBI circular Q&A, offshore PII export, high-value transfer, prompt injection and more), submit it, and watch guardrails intercept the action step by step.

    • Each run shows parsing, context, signals, guardrail hits and the final decision.
    • The live report and activity feed update in real time as runs land.
    Open Demo Agent

  4. Author policy as code

    In the Policy Engine you can add runtime guardrails or knowledge base policies, edit them visually or directly as YAML, then publish to the catalog.

    • Use the Visual tab for guided fields, the Code tab for raw YAML.
    • Copy or download the YAML to keep it in your own repo.
    Open Policy Engine

  5. Simulate before you enforce

    The simulation runner dry-runs a policy against sample audit decisions and tells you which controls would breach, near-miss or pass.

    • Turn on What-if mode to edit policy code and compare draft vs published.
    • Every run is archived under Run history with its full trace.
    • The Control heatmap summarises breach frequency by control.
    Open simulation runner

  6. Clear the oversight queue

    Human Approvals, Risk Management and the Audit Center are where governance work actually gets closed out.

    • Approvals show SLA targets and risk level per item.
    • Audit Center holds the tamper-evident decision ledger for evidence.
    Open Human Approvals

  7. Ask the CRO Copilot

    Finish with the CRO Copilot — a threaded assistant grounded in your risk register, approvals queue, policy catalog and decision ledger.

    • Threads are saved, so you can revisit a briefing later.
    • Answers cite the record IDs they were derived from.
    Open CRO Copilot