Observability on the Build Your Own AI Platform track. Log latency, errors, version, token counts, and a sample of failures. Trace a request id from the gateway to the model.
This lesson assumes you already worked through Quotas.
The idea in practice
Dashboards need a burn alert on error rate and latency. Keep enough detail to replay a bad request without storing secrets.
A concrete check
goal = {
'track': 'Build Your Own AI Platform',
'lesson': 'Observability',
}
checks = [
'input available at decision time',
'score matches the real decision',
'failure case written down',
]
print(goal['lesson'])
for item in checks:
print('-', item)
Run the sketch locally if you have Python. The printout is a reminder of the checks, not a trained model. Replace the strings with the real inputs from your own example before you treat it as a design.
What usually goes wrong
Logs so verbose they contain full customer documents forever. When this happens, stop adding parameters or tools. Fix the check, the data, or the permission, then run the same example again.
What to write down
- The input you are allowed to use at decision time.
- The output and the score or pass rule.
- One failure you will test on purpose.
- What you will not claim the system can do.
Practice
List five fields on the request log and two you redact.
Self-check
- Say Observability in one sentence that mentions an input and an output.
- Name the failure mode in this lesson and the check that would catch it.
Done when: you can explain this lesson without the page open, and you have a written failure case.