Build RAG Chatbot with Llamaindex, Pgvector, Mistral Ministral 3B, and Ollama granite-embedding

Introduction to RAG

Retrieval-Augmented Generation (RAG) is a game-changer for GenAI applications, especially in conversational AI. It combines the power of pre-trained large language models (LLMs) like OpenAI’s GPT with external knowledge sources stored in vector databases such as Milvus and Zilliz Cloud, allowing for more accurate, contextually relevant, and up-to-date response generation. A RAG pipeline usually consists of four basic components: a vector database, an embedding model, an LLM, and a framework.

Key Components We'll Use for This RAG Chatbot

This tutorial shows you how to build a simple RAG chatbot in Python using the following components:

Llamaindex: a data framework that connects large language models (LLMs) with various data sources, enabling efficient retrieval-augmented generation (RAG). It helps structure, index, and query private or external data, optimizing LLM applications for search, chatbots, and analytics.
Pgvector: an open-source extension for PostgreSQL that enables efficient storage and querying of high-dimensional vector data, essential for machine learning and AI applications. Designed to handle embeddings, it supports fast approximate nearest neighbor (ANN) searches using algorithms like HNSW and IVFFlat. Since it is just a vector search add-on to traditional search rather than a purpose-built vector database, it lacks scalability and availability and many other advanced features required by enterprise-level applications. Therefore, if you prefer a much more scalable solution or hate to manage your own infrastructure, we recommend using Zilliz Cloud, which is a fully managed vector database service built on the open-source Milvus and offers a free tier supporting up to 1 million vectors.)
Mistral Ministral 3B: A compact, high-efficiency language model optimized for fast inference and low-resource environments. With 3 billion parameters, it balances performance and scalability, excelling in text generation, summarization, and question-answering tasks. Ideal for edge computing, real-time applications, and cost-sensitive deployments requiring reliable NLP capabilities without heavy computational demands.
Ollama Granite-Embedding: A high-performance embedding model designed for semantic understanding and retrieval tasks. It excels at generating dense vector representations for text, enabling robust similarity search, clustering, and retrieval-augmented generation (RAG). Ideal for enterprise applications requiring scalable, privacy-preserving semantic analysis in on-premises or edge environments.

By the end of this tutorial, you’ll have a functional chatbot capable of answering questions based on a custom knowledge base.

Note: Since we may use proprietary models in our tutorials, make sure you have the required API key beforehand.

Step 1: Install and Set Up Llamaindex

pip install llama-index

Step 2: Install and Set Up Mistral Ministral 3B

%pip install llama-index-llms-mistralai

from llama_index.llms.mistralai import MistralAI

llm = MistralAI(model="ministral-3b-latest")

Step 3: Install and Set Up Ollama granite-embedding

%pip install llama-index-embeddings-ollama

from llama_index.embeddings.ollama import OllamaEmbedding

embed_model = OllamaEmbedding(
    model_name="granite-embedding",
)

Step 4: Install and Set Up Pgvector

%pip install llama-index-vector-stores-postgres

from llama_index.core import VectorStoreIndex
from llama_index.vector_stores.postgres import PGVectorStore

vector_store = PGVectorStore.from_params(
    database=db_name,
    host=url.host,
    password=url.password,
    port=url.port,
    user=url.username,
    table_name="your_table_name",
    embed_dim=1536,  # openai embedding dimension
    hnsw_kwargs={
        "hnsw_m": 16,
        "hnsw_ef_construction": 64,
        "hnsw_ef_search": 40,
        "hnsw_dist_method": "vector_cosine_ops",
    },
)

Step 5: Build a RAG Chatbot

Now that you’ve set up all components, let’s start to build a simple chatbot. We’ll use the Milvus introduction doc as a private knowledge base. You can replace it with your own dataset to customize your RAG chatbot.

import requests

from llama_index.core import SimpleDirectoryReader

# load documents
url = 'https://raw.githubusercontent.com/milvus-io/milvus-docs/refs/heads/v2.5.x/site/en/about/overview.md'
example_file = 'example_file.md'  # You can replace it with your own file paths.
response = requests.get(url)
with open(example_file, 'wb') as f:
    f.write(response.content)
documents = SimpleDirectoryReader(
    input_files=[example_file]
).load_data()

print("Document ID:", documents[0].doc_id)

storage_context = StorageContext.from_defaults(vector_store=vector_store)
index = VectorStoreIndex.from_documents(
    documents, storage_context=storage_context, embed_model=embed_model
)

query_engine = index.as_query_engine(llm=llm)
res = query_engine.query("What is Milvus?")  # You can replace it with your own question.
print(res)

Example output

Milvus is a high-performance, highly scalable vector database designed to operate efficiently across various environments, from personal laptops to large-scale distributed systems. It is available as both open-source software and a cloud service. Milvus excels in managing unstructured data by converting it into numerical vectors through embeddings, which facilitates fast and scalable searches and analytics. The database supports a wide range of data types and offers robust data modeling capabilities, allowing users to organize their data effectively. Additionally, Milvus provides multiple deployment options, including a lightweight version for quick prototyping and a distributed version for handling massive data scales.

Optimization Tips

As you build your RAG system, optimization is key to ensuring peak performance and efficiency. While setting up the components is an essential first step, fine-tuning each one will help you create a solution that works even better and scales seamlessly. In this section, we’ll share some practical tips for optimizing all these components, giving you the edge to build smarter, faster, and more responsive RAG applications.

LlamaIndex optimization tips

To optimize LlamaIndex for a Retrieval-Augmented Generation (RAG) setup, structure your data efficiently using hierarchical indices like tree-based or keyword-table indices for faster retrieval. Use embeddings that align with your use case to improve search relevance. Fine-tune chunk sizes to balance context length and retrieval precision. Enable caching for frequently accessed queries to enhance performance. Optimize metadata filtering to reduce unnecessary search space and improve speed. If using vector databases, ensure indexing strategies align with your query patterns. Implement async processing to handle large-scale document ingestion efficiently. Regularly monitor query performance and adjust indexing parameters as needed for optimal results.

pgvector optimization tips

To optimize pgvector in a Retrieval-Augmented Generation (RAG) setup, consider indexing your vectors using GiST or IVFFlat to significantly speed up search queries and improve retrieval performance. Make sure to leverage parallelization for query execution, allowing multiple queries to be processed simultaneously, especially for large datasets. Optimize memory usage by tuning the vector storage size and using compressed embeddings where possible. To further enhance query speed, implement pre-filtering techniques to narrow down search space before querying. Regularly rebuild indexes to ensure they are up to date with any new data. Fine-tune vectorization models to reduce dimensionality without sacrificing accuracy, thus improving both storage efficiency and retrieval times. Finally, manage resource allocation carefully, utilizing horizontal scaling for larger datasets and offloading intensive operations to dedicated processing units to maintain responsiveness during high-traffic periods.

Mistral Ministral 3B optimization tips

To optimize Mistral Ministral 3B in RAG, fine-tune it on domain-specific data to improve retrieval relevance and response accuracy. Use 4-bit or 8-bit quantization to reduce memory usage while maintaining performance. Implement dynamic batching during inference to handle multiple queries efficiently. Prune redundant layers or apply LoRA for lightweight adaptation. Cache frequent retrieval outputs to minimize recomputation. Optimize prompt engineering to reduce input token length, and leverage FlashAttention for faster processing. Monitor latency and adjust context window sizes based on use-case requirements to balance speed and coherence.

Ollama Granite-Embedding optimization tips

To optimize Ollama Granite-Embedding in RAG, ensure input text is cleanly chunked (avoid truncation by splitting documents into 512-token segments). Fine-tune embedding parameters like temperature and batch size for speed-quality balance. Use hardware acceleration (e.g., CUDA) and quantize the model for faster inference. Normalize embeddings to improve similarity calculations. Regularly evaluate retrieval accuracy with benchmarks like NDCG or recall@k. Cache frequent queries to reduce redundant computations, and pre-filter low-relevance documents using metadata to lighten embedding workloads.

By implementing these tips across your components, you'll be able to enhance the performance and functionality of your RAG system, ensuring it’s optimized for both speed and accuracy. Keep testing, iterating, and refining your setup to stay ahead in the ever-evolving world of AI development.

RAG Cost Calculator: A Free Tool to Calculate Your Cost in Seconds

Estimating the cost of a Retrieval-Augmented Generation (RAG) pipeline involves analyzing expenses across vector storage, compute resources, and API usage. Key cost drivers include vector database queries, embedding generation, and LLM inference.

RAG Cost Calculator is a free tool that quickly estimates the cost of building a RAG pipeline, including chunking, embedding, vector storage/search, and LLM generation. It also helps you identify cost-saving opportunities and achieve up to 10x cost reduction on vector databases with the serverless option.

Calculate your RAG cost now.

Calculate your RAG cost

What Have You Learned?

Wow, what a journey you've just embarked on! After diving into this tutorial, you now have a robust understanding of how to seamlessly integrate cutting-edge technologies to build an effective Retrieval-Augmented Generation (RAG) system. By harnessing the power of LlamaIndex as your framework, combined with Pgvector for efficient vector data storage, and leveraging both Mistral Ministral 3B as your Large Language Model (LLM) and Ollama granite-embedding for sophisticated embeddings, you’ve learned how each component plays a vital role in enriching your application's performance. Each piece not only fits together beautifully but also enhances the capabilities of your RAG pipeline, enabling you to pull in relevant information effortlessly and generate thoughtful, context-aware responses.

What’s even more exciting is the extra optimization tips and the handy free RAG cost calculator you picked up along the way! These insights will undoubtedly help streamline your development process, making your implementation even more efficient and budget-friendly. Now that you're armed with this knowledge, the sky's the limit! So go ahead, unleash your creativity, and start building, optimizing, and innovating your own RAG applications. Every great project begins with a single step; let this tutorial be your launchpad to creating something truly amazing. The world of RAG is waiting for your unique touch—get started today, and who knows what incredible solutions you'll bring to life!

Further Resources

🌟 In addition to this RAG tutorial, unleash your full potential with these incredible resources to level up your RAG skills.

How to Build a Multimodal RAG | Documentation
How to Enhance the Performance of Your RAG Pipeline
Graph RAG with Milvus | Documentation
How to Evaluate RAG Applications - Zilliz Learn
Generative AI Resource Hub | Zilliz

We'd Love to Hear What You Think!

We’d love to hear your thoughts! 🌟 Leave your questions or comments below or join our vibrant Milvus Discord community to share your experiences, ask questions, or connect with thousands of AI enthusiasts. Your journey matters to us!

If you like this tutorial, show your support by giving our Milvus GitHub repo a star ⭐—it means the world to us and inspires us to keep creating! 💖