enterprise-search

What problem does it solve?

Enterprise search often ships with “it looks better” tweaks that fail in production, causing poor ranking, weak query understanding, and unmanaged index changes.

Core Features & Use Cases

• Mode-based enterprise search engineering across relevance tuning, query understanding, index management, quality measurement, and performance optimization.
• Metric-first tuning workflows that require baselines (e.g., nDCG@10, MRR, P@k) and isolated, one-change-at-a-time evaluation.
• LLM guardrails for search tasks to reduce hallucinated or version-wrong Query DSL and to enforce specificity, tradeoffs, and regression checks.

Quick Start

Use the enterprise-search skill to improve ranking for an OpenSearch or Elasticsearch use case by tuning analyzers and boosts after capturing baseline relevance metrics.