@cocreate/qdrant

qdrant

A highly efficient, enterprise-grade Qdrant vector database CRUD wrapper and state-routing backend client for the collaborative CoCreate ecosystem. This wrapper maps Qdrant's high-performance neural search, payload filtering, and vector similarity capabilities into CoCreate's unified JSON interface.

Documentation

For a complete guide and working integration examples, refer to the CoCreate Qdrant Documentation.

---

Table of Contents

• The Power of the CoCreate Unified Ecosystem
• Why Qdrant?
• Installation
• Unified JSON Payload Examples
• Announcements
• Roadmap
• How to Contribute
• About

---

The Power of the CoCreate Unified Ecosystem

This is not just another database driver. By wrapping Qdrant into the CoCreate state-routing pipeline, you unlock the full power of a multi-cloud, multi-model data orchestration mesh:

• Unified JSON Interface (MongoDB-like Syntax): You don't need to write custom gRPC/REST vector payloads or manually isolate geometric distances. CoCreate translates standard NoSQL JSON CRUD operations (such as object.create or object.read) transparently into optimized Qdrant point insertions and vector search calls under the hood.
• Parallel Multi-Database Writing: Write to many target databases simultaneously with a single API call. You can stream transactional data to MongoDB (for instant operational state), persist it in PostgreSQL (for enterprise relational analytics), and vectorize it into Qdrant (for real-time semantic search) in perfect parallel harmony:

$$\text{Write Command} \xrightarrow{\text{Parallel Mesh}} \begin{cases} \text{MongoDB} & \text{(ACID Transaction Store)} \ \text{PostgreSQL} & \text{(Enterprise Relational Core)} \ \text{Qdrant} & \text{(Semantic Vector Search)} \end{cases}$$

• Dynamic State-Routing: Isolate multi-tenant environments effortlessly. The routing mesh automatically injects tenant metadata and organization boundaries, ensuring isolated database processing within your shared Qdrant collection infrastructure.

---

Why Qdrant?

Qdrant is an open-source vector database and vector similarity search engine designed to handle high-dimensional embeddings generated by LLMs, computer vision models, and other neural networks. It features full payload filtering support, allowing you to filter semantic results using standard unstructured metadata attributes alongside geometric vector matches.

By routing machine learning inputs, semantic application profiles, and structural unstructured information straight to Qdrant via this wrapper, you leverage ultra-low-latency similarity search indexing while maintaining a zero-friction JSON pipeline.

---

Installation

NPM Installation

npm i @cocreate/qdrant

Yarn Installation

yarn add @cocreate/qdrant

---

Unified JSON Payload Examples

Here are the complete, real-world JSON objects representing raw payloads routed through the CoCreate system. By declaring an array in the storage key, CoCreate fires the write operations to all specified backends simultaneously in parallel.

1. Writing a Full Transaction Object (object.create)

When this operational payload is sent, CoCreate registers it, processes the nested JSON fields, and automatically warehouses it across all listed parallel storage platforms.

{
  "method": "object.create",
  "storage": ["qdrant", "mongodb", "postgresql"], // Array of storages to store in parallel!
  "database": "sales_analytics",
  "array": "order_events",
  "object": {
    "_id": "order_77c8f2a9_3b01",
    "customer": {
      "id": "cust_88301",
      "email": "customer@example.com",
      "region": "North America",
      "acquisition_channel": "Google Search"
    },
    "transaction": {
      "payment_method": "Credit Card",
      "currency": "USD",
      "subtotal": 249.99,
      "discount": 15.00,
      "tax": 18.80,
      "total": 253.79
    },
    "items": [
      {
        "product_id": "prod_head_009",
        "name": "Wireless Noise-Cancelling Headphones",
        "category": "Electronics",
        "quantity": 1,
        "unit_price": 199.99
      },
      {
        "product_id": "prod_stand_012",
        "name": "Aluminum Headphone Stand",
        "category": "Accessories",
        "quantity": 2,
        "unit_price": 25.00
      }
    ],
    "shipping": {
      "carrier": "FedEx",
      "service": "2-Day Air",
      "estimated_delivery": "2026-07-20T12:00:00.000Z"
    }
  },
  "organization_id": "org_enterprise_99x",
  "user_id": "usr_sales_runner_402",
  "timeStamp": "2026-07-18T18:40:43.000Z"
}

2. Reading and Filtering Warehoused Data (object.read)

Querying targets a single database engine via a string filter syntax. The query below retrieves high-value orders handled by a specific shipping carrier from Qdrant, sorting them by total sale volume.

{
  "method": "object.read",
  "storage": "qdrant",
  "database": "sales_analytics",
  "array": "order_events",
  "$filter": {
    "query": {
      "transaction.total": { "$gte": 100.00 },
      "shipping.carrier": "FedEx"
    },
    "sort": [
      { "key": "transaction.total", "direction": "desc" }
    ],
    "limit": 50,
    "index": 0
  },
  "organization_id": "org_enterprise_99x"
}

---

Announcements

All updates to this library are documented in our CHANGELOG and releases. You may also subscribe to email for releases and breaking changes.

---

Roadmap

If you are interested in the future direction of this project, please take a look at our open issues and pull requests. We would love to hear your feedback.

---

How to Contribute

We encourage contribution to our libraries (you might even score some nifty swag), please see our CONTRIBUTING guide for details.

We want this library to be community-driven, and CoCreate led. We need your help to realize this goal. To help make sure we are building the right things in the right order, we ask that you create issues and pull requests or merely upvote or comment on existing issues or pull requests.

We appreciate your continued support, thank you!

---

About

CoCreate-qdrant is guided and supported by the CoCreate Developer Experience Team.

Please Email the Developer Experience Team here in case of any queries.

CoCreate-qdrant is maintained and funded by CoCreate. The names and logos for CoCreate are trademarks of CoCreate, LLC.