Question 3Integrations and APIs
A customer's API starts returning errors in production at 2am. What do you do?
FDEs often own production for their customers. This tests incident handling and communication.
“Hands-on with APIs, webhooks, authentication (OAuth, SSO) and data integrations”Forward Deployed Engineer · Harmony
The answerFour steps
- 1
Check your side first: logs, recent deploys, retries piling up
Say it like this
“First fifteen minutes: I check our side before blaming theirs.”
Why it scores
Our own change is often the real cause. Checking it first also shows you don't point fingers.
- 2
Stop the damage (pause jobs, queue requests, turn off the feature)
Say it like this
“I pause the sync and queue the requests, so nothing is lost and we don't hammer them while they recover.”
Why it scores
Contains the damage without losing data. They want mitigation before diagnosis.
- 3
Tell the customer what you see and what you've done, in plain words
Say it like this
“Then I message their on-call contact and our account owner: what broke, since when, what I've done, and when they'll hear from me next.”
Why it scores
A status update with a time for the next one. This is what the customer remembers.
- 4
Afterwards: add a check that would have caught it, and write a short report
Say it like this
“an alert that fires when their error rate rises, before their users notice, and a one-page write-up with the timeline.”
Why it scores
Closes the loop. Without this part the answer reads as firefighting only.
StrongAll together
About a minute, out loud. The marked parts are the four steps; hover one to find it above.
First fifteen minutes: I check our side before blaming theirs.1 Did we deploy anything? Are our retries piling up and making it worse? If their API really is down, I pause the sync and queue the requests, so nothing is lost and we don't hammer them while they recover.2 Then I message their on-call contact and our account owner: what broke, since when, what I've done, and when they'll hear from me next.3 Short, no guesses. Next day: an alert that fires when their error rate rises, before their users notice, and a one-page write-up with the timeline.4
WeakThe answer that fails
Sounds fine. Common. Scores low.
“I'd open the logs and debug until I find the root cause, then deploy a fix and let the customer know it's resolved.”
Interviewer's note
Silent for hours while the customer's data is broken. No mitigation, no update until the end.
PracticeTwo minutes, then they push back
Your turn
A customer's API starts returning errors in production at 2am. What do you do?
Answer like you're in the room: about two minutes, no notes. Say it out loud, even alone. It feels different from thinking it. Then check your answer against what interviewers score.