Using semantic benchmarks to build a self-improving text-to-query agentconversion.ai3 points·jimmyl02··0 commentsOpen articleSaveView on HN