Transformations and KPIs with AI Tools | Dataford Interview Questions

Given an existing code base, implement the data transformation code and generate the possible KPIs, using AI tools and explaining how you use them

Asked in the AI coding round stage. Full AI access; evaluation includes both the implementation and how effectively you use and verify AI assistance.

Implement transform_and_generate_kpis(records). Each record is a dictionary with timestamp as an ISO string, user_id, event, and numeric value. Return {"records": transformed_records, "kpis": kpis}. Add day from the first 10 timestamp characters and normalize event with trimming and lowercase. KPIs must include total events, unique users, total value, average value rounded to two decimals, counts by event, and counts by day.

Constraints

0 <= len(records) <= 5000
Each record contains timestamp, user_id, event, and value
timestamp is a non-empty ISO-formatted string beginning with YYYY-MM-DD
event is a string and value is numeric
Preserve the input record order

Function Signature

def transform_and_generate_kpis(records):

Your question is Transformations and KPIs with AI Tools. Start with the requirements 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

Given an existing code base, implement the data transformation code and generate the possible KPIs, using AI tools and explaining how you use them

Asked in the AI coding round stage. Full AI access; evaluation includes both the implementation and how effectively you use and verify AI assistance.

Constraints

0 <= len(records) <= 5000
Each record contains timestamp, user_id, event, and value
timestamp is a non-empty ISO-formatted string beginning with YYYY-MM-DD
event is a string and value is numeric
Preserve the input record order

Function Signature

def transform_and_generate_kpis(records):

Your question is Transformations and KPIs with AI Tools. Start with the requirements 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.