Add YugabyteDB as a Vector Store

[Sfurti-yb]

Description

Hi, I am reaching out on behalf of YugabyteDB,a high-performance, PostgreSQL-compatible distributed database built for global, internet-scale applications.

This contribution adds support for YugabyteDB as a vector store in LlamaIndex. The implementation is based on the existing PostgreSQL vector store, with modifications and enhancements specific to YugabyteDB’s architecture.

We’d appreciate it if you could review the changes and guide us on the next steps to have YugabyteDB officially listed as a supported vector store in LlamaIndex—similar to the Postgres example.

Looking forward to your feedback!

Fixes # (issue)

New Package?

Did I fill in the tool.llamahub section in the pyproject.toml and provide a detailed README.md for my new integration or package?

• Yes
• No

Version Bump?

Did I bump the version in the pyproject.toml file of the package I am updating? (Except for the llama-index-core package)

• Yes
• No

Type of Change

• New feature (non-breaking change which adds functionality)
• This change requires a documentation update

How Has This Been Tested?

Your pull-request will likely not be merged unless it is covered by some form of impactful unit testing.

• I added new unit tests to cover this change

Suggested Checklist:

• I have performed a self-review of my own code
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• I have added Google Colab support for the newly added notebooks.
• My changes generate no new warnings
• I have added tests that prove my fix is effective or that my feature works
• New and existing unit tests pass locally with my changes
• I ran uv run make format; uv run make lint to appease the lint gods

[Sfurti-yb]

@logan-markewich Extending the PostgresVectorStore currently introduces psycopg2-binary as a transitive dependency. This creates a conflict with psycopg2-yugabytedb, our smart driver, which provides features such as load balancing to better leverage YugabyteDB’s distributed architecture.

Is there a way to inherit PGVectorStore without bringing in psycopg2-binary as a transitive dependency?

If this is not feasible, we propose proceeding with the current implementation without inheriting PGVectorStore. However, if it is important for your use case that we directly inherit the PGVectorStore code, we can make the necessary changes—though this would require removing certain smart driver features like load balancing.

Please let us know your preference so we can align accordingly.

[Sfurti-yb]

@logan-markewich Gentle reminder here. Please let us know if there is something on our end that needs to be done. Thanks

[Sfurti-yb]

@logan-markewich Gentle reminder to take a look at the PR. Thanks.

[Sfurti-yb]

@AstraBert Could you also take a look at this PR. Thanks.