← AT&T Interview Insights

AT&T·Software Engineer·Technical Phone Screen·Intermediate

Intermediate
May 2026

Summary

Short technical screen for a data engineer role, basically just one SQL question and that was it. Pretty low-key, nothing fancy.

Questions Asked (1)

Q1

You have two tables: one containing streamer information and one containing streaming session records. Write a query to compute the total view count for each streaming session, grouped by streamer.

Data ModelingProduct Analytics & Metrics
Author's notes

Pretty standard join question.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying the schema: identify the join key between the streamer and session tables, and confirm that each session has a view count. Then write a query that joins the tables, groups by streamer and session, and sums the view count. Finally, discuss edge cases like sessions with no views or streamers with no sessions.

Pro tip: Mention that you would validate the grain of the session table (e.g., one row per session) to avoid double-counting views, and consider using a LEFT JOIN to include streamers with zero sessions if the business needs a complete list.

1. Clarify requirements and schema

Ask about the table structures, join keys, and whether the output should include streamers with no sessions. Confirm the definition of 'total view count' (e.g., sum of views per session).

2. Identify join and grouping keys

Determine the common column (e.g., streamer_id) to join the tables. Decide whether to group by streamer and session, or just by streamer if session-level detail is not needed.

3. Write the SQL query

Construct a query using JOIN, GROUP BY, and SUM. Use aliases for readability and consider using COALESCE to handle NULLs if using LEFT JOIN.

4. Validate and optimize

Check for duplicate sessions or multiple view records per session. Discuss indexing on join and group by columns for performance.

5. Discuss edge cases and extensions

Address scenarios like streamers with no sessions, sessions with zero views, and how to extend the query for additional metrics (e.g., average views per session).

Key Points to Mention

  • Use of INNER JOIN vs LEFT JOIN depending on whether to include streamers with no sessions
  • Grouping by streamer and session to compute total views per session
  • Aggregation function SUM on the view count column
  • Handling NULL values with COALESCE or IFNULL
  • Performance considerations: indexing on join keys and grouping columns
  • Data validation: ensuring no duplicate session records that could inflate view counts

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.