Building a Multi-Agent Self-Evolving System for ML Optimization

How we turned an OpenEvolve-style loop into a spec-driven, auditable framework — and ran it against a real two-tower retrieval pipeline Continue reading on Medium »