What problem does it solve? Qdrant deployments suffer from slow queries, sluggish indexing, and excessive RAM consumption, and finding the right fix requires matching a vague symptom to the correct tuning domain. This Skill acts as a navigation hub that maps user-reported symptoms to the appropriate sub-skill so the correct optimization guidance is applied instead of guesswork. ## Core Features & Use Cases - Symptom-Based Routing: Matches phrases like "filtered queries are slow" or "RAM keeps growing" to the correct sub-skill covering search speed, indexing performance, or memory usage. - Search Speed Optimization: Diagnoses latency and throughput problems, covering HNSW parameter tuning, quantization, batch search, payload indexes, and ACORN for filtered queries. - Indexing and Memory Tuning: Addresses slow bulk uploads, stuck optimizers, long HNSW builds, and high RAM usage via quantization, on-disk storage tiers, and MRL dimensionality reduction. - Use Case: A user reports that filtered vector search became slow after data growth. The hub routes to the search-speed sub-skill, which recommends creating a payload index on the filtered field and triggering re-indexing to build filterable subgraph links. ## Quick Start Ask the assistant to diagnose why your Qdrant filtered queries are much slower than unfiltered ones and recommend configuration changes.
