March 2026 monthly summary for elastic/elasticsearch focusing on reliability and usability improvements in ML and ESQL components. Key outcomes: ML Forecast Tests Stability improved by increasing test timeout to reduce flaky failures; CHANGE_POINT in ESQL now accepts arguments in any order, backed by tests and updated docs and changelog. These efforts reduce CI noise, improve user experience for ESQL, and strengthen ML feature confidence.

February 2026 monthly summary focusing on delivering flexible NDJSON parsing, robust nested data handling in TextStructure, and improved observability for inference memory usage, with strong emphasis on reliability and code quality.

January 2026 focused on ML reliability improvements in elastic/elasticsearch. Implemented guided remediation for ML node allocation errors by surfacing a concrete fix path and enabling xpack.ml.use_auto_machine_memory_percent when issues occur; updated documentation and error flows to surface this guidance to affected users. Result: faster remediation, reduced downtime, and improved user experience for ML workloads.