Roadmap

We're thrilled to present a comprehensive overview of the features and projects we're planning to develop. We're eager to engage with our community and users to shape the priority of these initiatives. We value your input, so don't hesitate to share your thoughts with us.

• [Launched 8/7/2023] JavaScript/TypeScript Library: A library to interact EpsillaDB with JavaScript/TypeScript.

• CLI with SQL-style Syntax: A new command-line interface is in the works, featuring a syntax similar to SQL for intuitive, straightforward commands.

• Integration Support:

  • [Launched 8/23/2023] LangChain

  • [Launched 9/6/2023] LlamaIndex

  • OpenAI plugin

  • AutoGPT

• Hybrid Search Support: We plan to support hybrid search allows you to perform semantic and keyword search and combine the results for improving result relevance.

  • [Launched 12/04/2023] Support sparse vector

  • [Launched 2/15/2024] Hybrid search API

  • Integration with BM25 encoding

  • Integration with SPLADE embedding

• Built-in Embedding Support: We aim to streamline your workflow by enabling document embedding directly within Epsilla vector DB, eliminating the need for pre-insertion embedding.

• Multi-modal Document Support: EpsillaDB is expanding to support a variety of document formats, including images, audio, video, and more.

  • Support built-in embedding models for images, audios, and videos

  • Support built-in OpenClip embedding cross image and text modality

  • Support natively store images, audios, and videos

• Comprehensive DBMS Features: We're committed to enhancing our database management system features to provide you with a powerful, flexible tool for managing your data.

  • [Launched 8/20/2023] Primary key support

  • [Launched 9/11/2023] Delete by primary key

  • [Launched 9/15/2023] Filter expression in vector similarity search

  • [Launched 12/25/2023] Filter expression in delete

  • Batch query support: batch multiple input vectors into single query

  • [Launched 1/6/2024] Update/Upsert support

  • DB import/export support

• Serverless Cloud Offering: A consumption-based pricing model is on the horizon, bringing you the flexibility of a serverless cloud solution akin to DynamoDB for vector search.

  • [Launched 9/20/2023] Closed beta

  • [Launched 12/13/2023] Public beta

  • General availability

• Enterprise Readiness Features: We're implementing robust enterprise readiness features, such as a distributed vector engine, high availability, access control, integrated authentication and authorization, and observability, amongst others.

  • [Launched 11/21/2023] Distributed vector engine

  • [Launched 11/21/2023] High availability

• Embedding Visualization: To give you more insight into your data, we're working on a feature that visualizes embeddings directly within EpsillaDB.

Your insights are integral to our development journey. We're looking forward to hearing your thoughts on these proposed features and initiatives.