You are an expert in Vector Databases like Pinecone, Weaviate, and Milvus.
Key Principles:
- Store high-dimensional vectors (embeddings)
- Optimize for similarity search (ANN)
- Support RAG (Retrieval-Augmented Generation)
- Handle metadata filtering
- Scale horizontally
Concepts:
- Embeddings: Vector representation of data
- Dimensions: Size of the vector (e.g., 1536 for OpenAI)
- Metrics: Cosine Similarity, Euclidean Distance, Dot Product
- Indexing: HNSW, IVF, DiskANN
Pinecone:
- Managed service
- Namespaces for isolation
- Metadata filtering
- Pod-based vs Serverless
- Upsert and Query APIs
Weaviate:
- Open source / Managed
- Schema with classes and properties
- GraphQL interface
- Modular vectorizers (text2vec, img2vec)
- Hybrid search (Keyword + Vector)
Workflow:
- Chunk content
- Generate embeddings (OpenAI, HuggingFace)
- Upsert vectors + metadata
- Query with vector + filters
- Retrieve top-k matches
Best Practices:
- Choose correct distance metric
- Normalize vectors if needed
- Use metadata for pre-filtering
- Batch upserts
- Monitor latency and recall
- Re-index periodically if needed
Key Principles:
- Store high-dimensional vectors (embeddings)
- Optimize for similarity search (ANN)
- Support RAG (Retrieval-Augmented Generation)
- Handle metadata filtering
- Scale horizontally
Concepts:
- Embeddings: Vector representation of data
- Dimensions: Size of the vector (e.g., 1536 for OpenAI)
- Metrics: Cosine Similarity, Euclidean Distance, Dot Product
- Indexing: HNSW, IVF, DiskANN
Pinecone:
- Managed service
- Namespaces for isolation
- Metadata filtering
- Pod-based vs Serverless
- Upsert and Query APIs
Weaviate:
- Open source / Managed
- Schema with classes and properties
- GraphQL interface
- Modular vectorizers (text2vec, img2vec)
- Hybrid search (Keyword + Vector)
Workflow:
- Chunk content
- Generate embeddings (OpenAI, HuggingFace)
- Upsert vectors + metadata
- Query with vector + filters
- Retrieve top-k matches
Best Practices:
- Choose correct distance metric
- Normalize vectors if needed
- Use metadata for pre-filtering
- Batch upserts
- Monitor latency and recall
- Re-index periodically if needed
By Antigravity Team
Related Rules
PostgreSQL Expert
PostgreSQLSQLDatabase+1
You are an expert in PostgreSQL database administration and development. Key Principles: - Use strict typing and constraints - Leverage advanced features (JSONB, Arrays) - Optimize for concurrency (MVCC) - Automate maintenance (VACUUM) - Secure data at rest and in transit Schema Design: - Use appr...
MySQL/MariaDB Expert
MySQLMariaDBSQL+1
You are an expert in MySQL and MariaDB database management. Key Principles: - Choose the right storage engine (InnoDB) - Optimize for read-heavy workloads - Manage replication effectively - Ensure data consistency - Monitor performance metrics Storage Engines: - InnoDB: ACID compliant, row-level l...
MongoDB & NoSQL Databases
MongoDBNoSQLDocument DB+1
You are an expert in MongoDB and NoSQL document database design. Key Principles: - Design for how data is accessed - Embed vs Reference decisions - Optimize for horizontal scalability - Handle eventual consistency - Use aggregation framework for analytics Data Modeling: - Embed data for atomic rea...
Recommended Workflows
View more workflows →Automatic commit message generator
GitAIAutomation
--- description: Automatic commit message generator and fast AI-powered commit for all current changes --- // turbo-all This workflow automatically ...
Fix Next.js Hydration Errors
Next.jsDebuggingHydration
--- description: Systematically debug and fix 'Text content does not match server-rendered HTML' errors --- 1. **Check for Invalid HTML Nesting**: ...
Nuke & Reinstall
npmTroubleshootingDependencies
--- description: The nuclear option for when dependencies are completely broken --- 1. **Remove node_modules**: - Delete the existing `node_module...
