vector-index-tuning - Tune vector indexes for production performance
Optimize vector index latency, recall, memory usage, quantization, and scaling through parameter tuning and benchmarking.
Tags
Updated: 2026-10-03Capabilities
Typical Inputs
Typical Outputs
What this skill does
- Select index types
- Tune HNSW parameters
- Benchmark index performance
- Apply vector quantization
- Estimate memory usage
- Recommend HNSW settings
Inputs
- Vector datasets
- Query vectors
- Ground-truth results
- Recall targets
- Latency limits
- Memory limits
- Vector dimensions
- Index settings
Outputs
- Benchmark results
- HNSW recommendations
- Quantized vectors
- Quantization parameters
- Memory estimates
Requirements
- Python environment
- NumPy
- hnswlib for HNSW benchmarking
- scikit-learn for product quantization