It's Tuesday, 8:40am. You started as Ledgerline's first AI Ops Analyst forty minutes ago. Ledgerline is a Series-B fintech: 40,000 paying businesses, one overworked data team, and a wall-mounted dashboard the whole company glances at like a weather report. This morning the dashboard is green. Activation is up, churn is flat, the on-call agent reports 'no open incidents.' Everyone is calm. Then a support lead Slacks you: 'three enterprise customers churned last week and nobody flagged it — why does the board deck say churn is flat?' You have until Thursday to find out whether the dashboard is right and the support lead is panicking, or the dashboard is lying and the company is flying blind. In the next hour you'll learn to interrogate a metric, read an agent's triage decisions, catch the difference between activity and signal, and wire a check that makes the numbers honest.
You'll walk out able to
- Tell the difference between a metric that's healthy and one that's quietly broken
- Read an automated triage agent's decisions and spot when it's closing real incidents
- Find the definition bug behind a flat line that should be moving
- Write the verification cases that prove a metric is honest before you trust it
- Pick the dashboard that tells the truth over the one that looks impressive
Your first task
Is the triage agent helping?
Ledgerline runs an on-call agent that reads every alert from the data pipeline and decides: real incident, or noise. It closed 19 alerts overnight and paged no one. The support lead's churn complaint made you nervous, so you pull the agent's overnight session and start reading.
What you'll answer
Watch the agent triage the overnight alerts below. Decide what to do about this agent right now.
Free · no card · we save your progress
Save your progress & earn the certificate
You've seen the opening — keep going free. 6 tasks · auto-graded · finish in 60 min and the certificate is yours. Free forever · no card · cancel any time.
We'll also email you a magic link so you can sign in on your phone. No password, ever.
Already started? Sign in with magic link instead.