skills.rest
google
googleOFFICIAL
Explore all skills by this author →
Website𝕏@GoogleOSSGitHub

google-cloud-solution-hybrid-search-alloydb

Generates architecture, design, and deployment guidance for AlloyDB hybrid search solutions on Google Cloud.

19.1k|1.5k|Updated Mar 31, 2026
Search Tags:#Google Cloud#solution architecture#cloud run#scann#Vector search#AlloyDB#Hybrid Search
One-click install
npx skills add https://github.com/google/skills --skill google-cloud-solution-hybrid-search-alloydb
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: google-cloud-solution-hybrid-search-alloydb
Source: https://github.com/google/skills/tree/main/skills/cloud/google-cloud-solution-hybrid-search-alloydb
Command: npx skills add https://github.com/google/skills --skill google-cloud-solution-hybrid-search-alloydb

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) and assets (resource) components.

What problem does it solve?

Designing a hybrid search system that combines vector similarity with structured SQL filtering requires coordinating many Google Cloud products, and teams often struggle to choose the right database, embedding engine, and hosting topology while meeting security, performance, and cost requirements.

Core Features & Use Cases

  • Guided Requirements Discovery: Walks through functional and non-functional requirements (security, reliability, cost, operations, performance, sustainability) and resolves contradictions before any architecture is proposed.
  • Grounded Architecture Generation: Produces Mermaid architecture diagrams, product mappings, and design recommendations grounded in curated reference files covering AlloyDB ScaNN indexing, BigQuery vector search, MCP Toolbox, and Cloud Run.
  • Deployment and Validation Guidance: Generates AlloyDB DDL/SQL scripts, Terraform or gcloud commands, pre-deployment dry-run checks, and runtime verification steps, then packages everything into a solution architecture guide.
  • Use Case: A retail team needs semantic product search with faceted filters (category, color, price) at under 100ms latency. The skill gathers requirements, recommends AlloyDB with ScaNN indexes and in-database reranking, and produces deployable SQL and Cloud Run configuration.

Quick Start

Ask the assistant to design a hybrid search solution on Google Cloud using AlloyDB for your product catalog with vector search and faceted filtering.

Frequently Asked Questions about google-cloud-solution-hybrid-search-alloydb

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I build a hybrid search system on Google Cloud with AlloyDB?▼

Use this skill's four-phase workflow: gather functional and non-functional requirements, generate a product mapping and Mermaid architecture diagram, produce design recommendations and deployment scripts, then validate with dry-run checks. It recommends AlloyDB with ScaNN vector indexes combined with SQL WHERE filters for hybrid queries.

AlloyDB vs BigQuery for vector search, which should I choose?▼

AlloyDB suits low-latency transactional hybrid search with in-database embeddings, ai.rank reranking, and ScaNN indexing. BigQuery vector search fits petabyte-scale analytical RAG and batch embedding generation. The skill's product mapping reference details trade-offs for both options.

How do I tune ScaNN index parameters in AlloyDB for better recall?▼

Set num_leaves to sqrt(rows) for a two-level tree or power(rows, 2/3) for three-level trees, and use scann.satisfy_limit = 'relaxed_order' with scann.max_pct_leaves_to_search = 15 for selective filters. The design recommendations reference covers tuning for high-dimensional embeddings.

When should I not use this hybrid search skill?▼

Avoid it for simple keyword-only search workloads or when a standalone non-relational vector database is required. It is optimized for cases combining vector search with structured SQL filtering, faceted attributes, and semantic reranking on managed relational databases.

Does the skill support deploying the search application on Cloud Run?▼

Yes, it generates deployment guidance including MCP Toolbox configuration on Cloud Run, Cloud Run Function shims, gcloud run deploy commands, and Terraform infrastructure code. It also covers Direct VPC Egress for private connectivity to AlloyDB.

Related Skills

Explore More Software Engineering Skills→
drgaciwdrgaciw

hybrid-cloud-networking

Configure VPNs and dedicated connections between on-premises networks and cloud platforms.

Community
Advanced
interflownexinterflownex
2

cloud-sql-postgres-vectorassist

Configure vector indexes and generate optimized SQL queries for Cloud SQL for PostgreSQL.

Community
Advanced
NihalmaiNihalmai
2

aws-well-architected-framework

Assesses AWS architecture against the six pillars of Well-Architected Framework.

Community
Advanced
cloudlinesscloudliness

pinecone

Host a scalable managed vector database for similarity search.

Community
Advanced