Kafka

• Fast is the easy part. Here is why you can trust KPipe with your offsets.
• Benchmarking KPipe against the parallel-Kafka libraries you would actually pick
• Kafka consumers get messy fast. KPipe is a simpler way to structure them.