Milvus DB: AI-Ready Vector Database Environment

This virtual machine bundles Milvus, the industry-leading open-source vector database, in a fully inte-grated environment designed for building and testing AI Agents with semantic search, and Retrieval-Augmented Generation (RAG) capabilities.

Ideal for developers, data scientists, and AI researchers, this VM offers a secure, private workspace with all the tools needed to work with vector embeddings, local language models, and interactive data exploration.

Milvus is an open-source, high-performance vector database built to accelerate applications involving unstructured data such as text, images, audio, and video. It's designed with both speed and scalability in mind, making it a preferred choice for modern AI, search, and recommendation systems.

**Key Features of Milvus:**

• **High-performance vector similarity search** (supports billion-scale data)
• **Multiple distance metrics** (L2, Cosine, Inner Product)
• **Hybrid search support** (combine vector and structured fields)
• **Scalable indexing options** (IVF, HNSW, etc.)
• **gRPC and RESTful APIs**

**Common Use Cases:**

• Semantic search engines
• Retrieval-Augmented Generation **(RAG)** pipelines
• Recommendation systems
• Visual similarity search (images, video, audio)
• Anomaly detection using embeddings

**Included Tools & Add-ons**

In addition to Milvus, this VM includes a curated set of tools to make development and experimenta-tion seamless:

**JupyterHub (with Python Virtual Environment)**

JupyterHub provides a multi-user, browser-based interface for running Jupyter notebooks. It enables interactive coding, data visualization, and experimentation in a shared Python environment.

• Accessible through the browser

**Pre-configured with:**

• pymilvus: Milvus Python SDK
• milvus-lite: lightweight in-memory version for testing
• ollama Python client

Provides a ready-to-run **RAG demo notebook** including:

• Document loading and embedding
• Vector insertion and search in Milvus
• Local LLM-based question answering

**Milvus CLI**

• Lightweight command-line tool for managing collections, indexes, and inspecting schemas
• Can be used as an alternative to WebUI for users who prefer terminal access

**Milvus Web UI**

• GUI for managing collections, viewing schema, and monitoring the database
• Restricted to RDP for security, as WebUI currently lacks authentication but can be made accessible through brows-er with ready to run script.

**Ollama LLM Runtime**

Ollama is a lightweight, local runtime for deploying and running large language models (LLMs) on your machine. It allows you to generate text, create embeddings, and build AI workflows without relying on external APIs.

• Supports embedding and generation models for local inference.
• Integrates with the RAG pipeline in the demo notebook.

**What's Included**

**Milvus (Docker):** Vector DB running in standalone mode

**JupyterHub:** Python IDE preloaded with SDKs & demo

**Milvus CLI:** Optional command-line tool for DB operations

**Ollama (host):** Local LLM runtime for embedding + generation

**Demo Notebook:** End-to-end RAG example pre-run and validated

**Ideal For**

• AI/ML engineers building GenAI apps or semantic search systems
• Researchers evaluating vector DBs and RAG architectures
• Teams building domain-specific search or retrieval tools
• Educational demos or internal POCs

**Secure & Private**

• All tools run locally inside the VM.
• Milvus Web UI is restricted to RDP for controlled access with the option to make it accessible in browser.
• Suitable for air-gapped or sensitive environments.

**Disclaimer:** Other trademarks and trade names may be used in this document to refer to either the entities claiming the marks and/or names or their products and are the property of their respective owners. We disclaim proprietary interest in the marks and names of others.