February 2020 Summaries

In February 2020, Vespa announced several product updates, including support for LightGBM machine learning models, which offer fast training, a small memory footprint, and improved accuracy. The update also highlighted the use of OpenBLAS for enhanced matrix multiplication performance within machine-learned models and introduced a benchmarking guide to optimize cost and performance for applications with strict latency requirements. Additionally, Vespa now features a fluent library for composing queries, thanks to contributions from the community, and has integrated with Hadoop, supporting conditional writes. Vespa, largely developed by Yahoo engineers, is an open-source big data processing and serving engine used by various Yahoo and Verizon Media products, and it continues to evolve with community feedback and contributions.