Dashboard for online games
Project idea
Build a real-time anomaly detection pipeline and a dashboard to show the aggregated results for users.
What is the purpose, and most common use cases?
Purpose:
- real-time anomaly detection based on user reporting and in-game aggregated stats
- dashboard that shows the most reported users
Use cases:
- detect game cheaters
Which technologies are well-suited?
Datasets:
- PUBG match deaths and statistics
- Simulated reporting events
Technologies:
- Kafka
- Flink
- Postgres
- Dash
What are the engineering challenges
- Stateful joining in Flink
Proposed architechture
Kafka -> Flink -> Postgres -> Dash