Build RAG Chatbot with Llamaindex, Pgvector, Mixtral 8x7B, and jina-clip-v2

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.)
Mixtral 8x7B: A sparse mixture-of-experts (MoE) model with eight 7B parameter networks, designed for efficient, high-performance NLP tasks. Excels in text generation, reasoning, and multilingual support while minimizing computational costs. Ideal for scalable enterprise applications, real-time chatbots, and multi-task environments requiring optimized resource utilization and versatile AI capabilities.
Jina-CLIP-V2: A multimodal AI model designed to seamlessly connect text and visual data, excelling in cross-modal retrieval tasks. Strengths include high accuracy in image-text matching, multilingual support, and scalable architecture. Ideal for semantic image search, content moderation, and personalized recommendations in e-commerce or media platforms.

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 Mixtral 8x7B

%pip install llama-index-llms-mistralai

from llama_index.llms.mistralai import MistralAI

llm = MistralAI(model="open-mixtral-8x7b")

Step 3: Install and Set Up jina-clip-v2

%pip install llama-index-embeddings-jinaai

You may also need other packages that do not come direcly with llama-index.

!pip install Pillow

from llama_index.embeddings.jinaai import JinaEmbedding

embed_model = JinaEmbedding(
    api_key=jinaai_api_key,
    model="jina-clip-v2",
    # choose `retrieval.passage` to get passage embeddings
    task="retrieval.passage",
)

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.

Mixtral 8x7B optimization tips

To optimize Mixtral 8x7B in RAG, prioritize efficient context retrieval by fine-tuning chunk size and overlap for balanced relevance and latency. Use sparse attention configurations to reduce computational overhead, and enable tensor parallelism to leverage its mixture-of-experts architecture. Quantize the model to 4-bit precision (e.g., via GPTQ) for faster inference with minimal accuracy loss. Pre-filter retrieved documents to remove noise, and cache frequent query embeddings. Adjust temperature (0.2-0.5) and max tokens to balance creativity and focus. Profile expert routing to ensure balanced workload distribution across GPU resources.

Jina-CLIP-v2 optimization tips

To optimize Jina-CLIP-v2 in a RAG setup, preprocess inputs by cleaning text, normalizing formats, and truncating overly long documents to reduce noise. Use batch inference for embeddings to leverage GPU parallelism, and ensure model weights are quantized or pruned for faster inference. Cache frequently accessed embeddings to avoid redundant computations. Fine-tune the model on domain-specific data to improve retrieval relevance. Pair with efficient vector indexes (e.g., FAISS) for low-latency similarity searches, and monitor embedding quality via recall metrics to balance speed and accuracy. Regularly update the index with fresh data.

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 an exciting journey you've just embarked on! In this tutorial, you’ve learned how to seamlessly integrate a powerful framework like LlamaIndex with a robust vector database such as Pgvector. Together, these components form the backbone of a Retrieval-Augmented Generation (RAG) system, allowing you to efficiently manage and retrieve data while harnessing the incredible capabilities of large language models (LLMs) like Mixtral 8x7B. You've seen how different parts of the RAG pipeline work hand-in-hand: the embedding model ensures that your vectors are optimized for search, while the LLM opens up endless possibilities for generating thoughtful responses or insights based on the retrieved information. This isn't just theory; you've gained the knowledge to implement these technologies together to create something truly impactful!

But we didn’t stop there! The tutorial also equipped you with some handy optimization tips to enhance the performance of your system, ensuring that it runs smoothly and efficiently, which is crucial for real-world applications. And let’s not forget about the free RAG cost calculator, a fantastic resource that helps you gauge the financial aspects of deploying your RAG applications. Now that you're armed with all this knowledge and these tools, it’s time to roll up your sleeves and start building, optimizing, and innovating your own incredible RAG applications! The world of AI is yours to explore, so dive in and make your mark!

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! 💖