SvectorDB Introduction
SvectorDB is a serverless vector database built for AWS, designed to offer cost-effective vector search solutions tailored for AWS environments. With a pay-per-use model, SvectorDB ensures that users only pay for what they need, optimizing cloud spend without any hidden costs. The platform simplifies the process of scaling products, allowing seamless transitions from prototype to production with minimal code.
SvectorDB Features
Scalability and Performance
SvectorDB handles the complexities of managing and scaling vector databases, enabling users to focus on their product's growth. Whether it's 1 vector or 1 million vectors, SvectorDB provides the necessary infrastructure to support seamless scaling.
# Create or update an item
client.setItem({
databaseId,
key: 'abc',
value: Buffer.from('Hello world!'),
vector: [0.1, 0.1, 0.1, 0.1]
});
# Query based on a vector
client.query({
databaseId,
query: {
vector: [0.5, 0.5, 0.5, 0.5]
}
});
# Query based on key (nearest to existing vector)
client.query({
databaseId,
query: {
key: 'abc'
}
});
Language Support and Integration
SvectorDB supports multiple programming languages, including JavaScript and Python, and offers an official OpenAPI specification for broader language compatibility. This flexibility makes it easy to integrate with various applications and workflows.
Hybrid Search
The platform allows users to perform Lucene/ElasticSearch style queries to filter results based on key-value pairs, enhancing the search capabilities and providing more granular control over the data.
Instant Updates
SvectorDB ensures that updates and deletions are reflected instantly, eliminating concerns about eventual consistency and providing real-time data accuracy.
Serverless Architecture
As a natively serverless solution, SvectorDB operates on a pay-per-request pricing model, eliminating the need for provisioning or scaling, and aligning costs with actual usage.
CloudFormation Support
SvectorDB can be easily integrated into existing CloudFormation templates, simplifying deployment and management within AWS environments.
Built-in Vectorizers
The platform comes with built-in vectorizers for text and images, allowing users to transform their data into vectors without the need for external tools. Additionally, users can bring their own embeddings to leverage the full power of SvectorDB.
client.embed({
model: EmbeddingModel.ALL_MINILM_L6_V2,
input: {
text: 'The quick brown fox jumps over the lazy dog.',
},
});
client.embed({
model: EmbeddingModel.CLIP_VIT_BASE_PATH32,
input: {
image: Buffer.from(...)
},
});
Pricing Transparency
SvectorDB offers transparent pricing with no minimum fees or upfront costs. Users only pay for the requests they make, ensuring cost control and predictability.
SvectorDB Pricing
Cost Breakdown
- Storage: $0.25/GB/month
- Queries: $5/million
- Writes: $20/million
Free Tier
SvectorDB provides a free tier with up to 5k records, allowing users to create and manage indexes without any time limits.
Price Comparison
When compared to other services like Pinecone, SvectorDB offers significant cost savings:
- SvectorDB: $7.25/month
- Pinecone: $83.44/month
SvectorDB Use Cases
Recommendation Engines
By representing items and users as vectors, recommendation engines can suggest relevant items to users based on their preferences and the similarity of vectors.
Document / Image Search
SvectorDB enables deep and meaningful search capabilities by transforming documents and images into vectors, allowing for semantic and visual similarity searches.
Retrieval Augmented Generation
Generative models can be augmented with context from SvectorDB, enhancing the quality of generated content and ensuring it is more refined and contextually relevant.
SvectorDB FAQs
What are the limitations of SvectorDB?
While SvectorDB offers robust features, it does have some limitations. These include the lack of snapshot capabilities, a default limit of 1 million records per database, and the size of the company, which may be a concern for some customers.
How does SvectorDB ensure data consistency?
SvectorDB maintains internal backups to protect against data loss, and updates are reflected instantly, ensuring data consistency without the need to worry about eventual consistency issues.
Can I use my own embeddings with SvectorDB?
Yes, SvectorDB allows users to bring their own embeddings or use the built-in vectorizers for text and images to leverage the full power of the platform.
