The first Kafka Summit conference took place on April 26, 2016. Confluent blogs summarized the highlights. Here are the talks I like most

• Leveraging Kafka for Big Data in Real Time Bidding, Analytics, ML & Campaign Management for Globally Distributed Data Flows by Helena Edelson.
• Streaming SQL by Julian Hyde.
• Securing Kafka by Jun Rao.
• Introducing Kafka Streams: Large-scale Stream Processing with Kafka by Neha Narkhede.

Now let's start our reading.

Streaming

• Last week, we saw Why Apache Beam? A Google Perspective. This week let's look at data Artisans perspective on their involvement in Apache Beam.

• In Streaming at Spark Summit East 2016, we talked about the upcoming Structured Streaming API in Spark 2.0. Now Spark 2.0 Technical Preview is available. They find a number of problems in adopting a single programming model integrating both batch and streaming data and

  The vision of Structured Streaming is to utilize the Catalyst optimizer to discover when it is possible to transparently turn a static program into an incremental execution that works on dynamic, infinite data (aka a stream).

Scala

• Li Haoyi explores how to make use of the type-safety of Scala to catch mistakes at compile time in Strategic Scala Style: Practical Type Safety. His previous posts in the series are highly recommended

  • Strategic Scala Style: Conciseness & Names
  • Strategic Scala Style: Principle of Least Power

• Daniel Westheide introduces his observation that

  People are using Option too often where their business logic clearly indicates they should use their own, custom ADT.

  in When Option Is Not Good Enough .

Machine Learning

• Google announced SyntaxNet, open sourcing the world’s most accurate parser. SyntaxNet is an open-source neural network framework implemented in TensorFlow that provides a foundation for Natural Language Understanding (NLU) systems.

One more thing

I also read about two nice articles on non-volatile memory this week

• NOVA: A Log-Structured File System for Hybrid Volatile/Non-Volatile Main Memories
• Non-volatile Storage from ACM Queue.

but I don't think I understand them well enough. Hence I'd like to defer them to a dedicated post. See you there.