Watch Time Completion Streaks

A content engagement analytics pipeline identifies highly engaged subscribers by combining a completion rate filter with a viewing streak detector, powering targeted retention campaigns. Given viewing session data, find all unique accounts that watched more than 80% of a title's runtime within the first 100 days from release. For qualifying accounts, return their top 3 titles that also meet a completion streak: watched on at least 3 different days within any 7-day window. **Input:** `sessions` (list of session dicts with user_id, title_id, watch_time, watch_date) and `titles` (list of title dicts with title_id, runtime, release_date). **Output:** Dict mapping qualifying user_ids to their top 3 streak titles. **Key Rules:** - 80% completion threshold applies within the 100-day post-release window only - Streak requires 3+ distinct days within any rolling 7-day window - Top 3 titles per qualifying account (by a defined ranking criterion)