Social Media Sentiment Tracking

Design a system that tracks public sentiment toward Netflix on social media over time: ingest posts, classify sentiment, aggregate, and surface trends so the team can see opinion shift around launches or price changes. The Challenge Design a system that tracks how people feel about Netflix on social media over time. The system must collect posts, classify sentiment as positive or negative, aggregate results, and expose trends so teams can see how public opinion changes around launches, pricing changes, or major announcements. Structure To Lead With Split the design into ingestion, aggregation, and offline ML. Ingestion collects posts from social platforms, aggregation buckets sentiment by time and topic, and the ML path trains or updates the sentiment classifier. Notes Clarify source coverage, rate limits, deduplication, and whether private or deleted content is out of scope. Define metrics beyond raw positive/negative counts: volume-normalized sentiment, moving averages, anomaly detection, and confidence bands. Treat the model as replaceable behind a scoring service; the system design should still work if the classifier improves over time.