High-Performance Java, Or How JVector Happened

airhacks.fm podcast with adam bien

An airhacks.fm conversation with Jonathan Ellis (@spyced) about:

Jonathan's first computer experiences with IBM PC 8086 and Thinkpad laptop with Red Hat Linux, becoming a key contributor to Apache Cassandra and founding datastax, starting DataStax to provide commercial support for Cassandra, early experiences with Java, C++, and python, discussion about the evolution of Java and its ecosystem, the importance of vector databases for semantic search and retrieval augmented generation, the development of JVector for high-performance vector search in Java, the potential of integrating JVector with LangChain for Java / langchain4j and quarkus for serverless deployment, the advantages of Java's productivity and performance for building concurrent data structures, the shift from locally installed software to cloud-based services, the challenges of being a manager and the benefits of taking a sabbatical to focus on creative pursuits, the importance of separating storage and compute in cloud databases, Cassandra's write-optimized architecture and improvements in read performance, DataStax's investment in Apache Pulsar for stream processing, the llama2java project for high-performance language models in Java

Jonathan Ellis on twitter: @spyced