Data Manipulation Gotchas | Dataford Interview Questions

Tell me about a time you handled tricky data manipulation issues, including edge cases and unexpected input, in a production system.

For a hands-on SQL follow-up, use the users and production_events tables. Treat supported event types as valid only when their timestamp is present and their value is either nonnegative or NULL.

Output

One row per user, including users with no events
Columns: user_id, email, valid_event_count, anomaly_count, and latest_valid_at
Exclude unknown users from the result, count invalid known-user events as anomalies, and sort by user_id ascending

Schema

users

Column	Type	Description
user_idPK	INT	Unique identifier for a known user
email	VARCHAR(255)	User email address
external_key	VARCHAR(64)	Optional identifier from an external system

production_events

Column	Type	Description
event_idPK	INT	Unique production event identifier
user_id	INT	User associated with the event, when known
event_type	VARCHAR(64)	Reported event type
occurred_at	TIMESTAMP	Time at which the event occurred
value	DECIMAL(10,2)	Optional nonnegative event value
source	VARCHAR(32)	Originating ingestion source

Your question is Data Manipulation Gotchas. Start with the requirements and the two tables in the Question tab.

Run and submit as often as you like. When you're ready, talk me through your approach or go straight to the code.

Problem