:rocket: Excited to share Part 3 of my "*Getting S...

# feed

[Jaehyeon Kim]

🚀 Excited to share Part 3 of my "*Getting Started with Real-Time Streaming in Kotlin*" series:

"*Kafka Streams - Lightweight Real-Time Processing for Supplier Stats*"!

After exploring Kafka clients with JSON and then Avro for data serialization, this post takes the next logical step into actual stream processing. We'll see how Kafka Streams offers a powerful way to build real-time analytical applications.

In this post, we'll cover:

• Consuming Avro order events for stateful aggregations.
• Implementing event-time processing using custom timestamp extractors.
• Handling late-arriving data with the Processor API.
• Calculating real-time supplier statistics (total price & count) in tumbling windows.
• Outputting results and late records, visualized with Kpow.
• Demonstrating the practical setup using Factor House Local and Kpow for a seamless Kafka development experience.
This is post 3 of 5, building our understanding before we look at Apache Flink. If you're interested in lightweight stream processing within your Kafka setup, I hope you find this useful!

Read the article: https://jaehyeon.me/blog/2025-06-03-kotlin-getting-started-kafka-streams/

Next, we'll explore Flink's DataStream API. As always, feedback is welcome!

🔗 Previous posts:
1. Kafka Clients with JSON
2. Kafka Clients with Avro