{
  "annotation_id": "be695647-8181-40b0-92d7-17ab034b797a",
  "date_created": "2026-02-17T05:53:07.827000Z",
  "date_modified": "2026-02-17T05:53:07.827000Z",
  "file_hash": "367cb283f90488a35e90f84ace5f1c051e5a57940ae9658ea31347bbc74f3692",
  "private": false,
  "record": {
    "abstract": "The rapid integration of vector search into AI applications, particularly for Retrieval Augmented Generation (RAG), has catalyzed the emergence of a diverse ecosystem of specialized vector databases. While this innovation offers a rich choice of features and performance characteristics, it has simultaneously introduced a significant challenge: severe API fragmentation. Developers face a landscape of disparate, proprietary, and often volatile API contracts, which hinders application portability, increases maintenance overhead, and leads to vendor lock-in. This paper introduces Vextra, a novel middleware abstraction layer designed to address this fragmentation. Vextra presents a unified, high-level API for core database operations, including data upsertion, similarity search, and metadata filtering. It employs a pluggable adapter architecture to translate these unified API calls into the native protocols of various backend databases. We argue that such an abstraction layer is a critical step towards maturing the vector database ecosystem, fostering interoperability, and enabling higher-level query optimization, while imposing minimal performance overhead.",
    "arxiv_id": "2601.06727",
    "authors": [
      "Chandan Suri",
      "Gursifath Bhasin"
    ],
    "categories": [
      "cs.DB",
      "cs.SE"
    ],
    "license": "http://arxiv.org/licenses/nonexclusive-distrib/1.0/",
    "title": "Vextra: A Unified Middleware Abstraction for Heterogeneous Vector Database Systems",
    "url": "https://arxiv.org/abs/2601.06727",
    "version": "v1"
  },
  "schema_id": "dorsal/arxiv",
  "source": {
    "execution_id": "36f87da5-ff0c-46c1-a693-3d4c507dfc4c",
    "id": "arXiv Dataset",
    "type": "Model",
    "variant": "snapshot-2026-01-17",
    "version": "0.1.0"
  },
  "user_id": 1000002
}