Build RAG Chatbot with Haystack, Milvus, Anthropic Claude 3 Sonnet, and BAAI bge-large-en-v1.5

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:

Haystack: An open-source Python framework designed for building production-ready NLP applications, particularly question answering and semantic search systems. Haystack excels at retrieving information from large document collections through its modular architecture that combines retrieval and reader components. Ideal for developers creating search applications, chatbots, and knowledge management systems that require efficient document processing and accurate information extraction from unstructured text.
Milvus: An open-source vector database optimized to store, index, and search large-scale vector embeddings efficiently, perfect for use cases like RAG, semantic search, and recommender systems. If you hate to manage your own infrastructure, we recommend using Zilliz Cloud, which is a fully managed vector database service built on Milvus and offers a free tier supporting up to 1 million vectors.
Anthropic Claude 3 Sonnet: A versatile AI model optimized for complex reasoning, multilingual tasks, and processing long-context inputs. It balances high performance with cost-efficiency, ideal for enterprise-scale applications like data analysis, real-time customer support, content creation, and research tasks requiring accuracy and scalability across diverse industries.
BAAI bge-large-en-v1.5: A dense embedding model optimized for semantic retrieval, excelling in capturing nuanced text semantics for tasks like retrieval-augmented generation (RAG), semantic search, and clustering. Its strengths include high accuracy across diverse domains, robust handling of complex queries, and efficient scalability. Ideal for enterprise search engines, recommendation systems, and knowledge-intensive NLP applications requiring precise contextual understanding.

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 Haystack

import os
import requests

from haystack import Pipeline
from haystack.components.converters import MarkdownToDocument

from haystack.components.preprocessors import DocumentSplitter
from haystack.components.writers import DocumentWriter

Step 2: Install and Set Up Anthropic Claude 3 Sonnet

To use Anthropic models, you need an Anthropic API key. You can provide this key in one of the following ways:

The recommended approach is to set it as the ANTHROPIC_API_KEY environment variable.
Alternatively, you can pass it directly when initializing the component using Haystack’s Secret API: Secret.from_token("your-api-key-here").

When configuring Anthropic models, make sure to define the Anthropic model you want to use by specifying it in the model parameter.

This component generates text based on a given prompt. Additionally, you can customize the generation process by providing extra parameters available in the Anthropic Messaging API. These parameters can be passed using generation_kwargs, either during initialization or when calling the run() method. To explore all available options, refer to the Anthropic documentation.

Finally, the run() method requires a single string as input to generate text.

Now let's install the anthropic-haystack package to use the AnthropicGenerator:

pip install anthropic-haystack

from haystack_integrations.components.generators.anthropic import AnthropicGenerator

generator = AnthropicGenerator(model="claude-3-sonnet-20240229")

Step 3: Install and Set Up BAAI bge-large-en-v1.5

from haystack import Document
from haystack.components.embedders import SentenceTransformersDocumentEmbedder
from haystack.components.embedders import SentenceTransformersTextEmbedder

doc_embedder = SentenceTransformersDocumentEmbedder(model="BAAI/bge-large-en-v1.5")
doc_embedder.warm_up()
text_embedder = SentenceTransformersTextEmbedder(model="BAAI/bge-large-en-v1.5")
text_embedder.warm_up()

Step 4: Install and Set Up Milvus

pip install --upgrade pymilvus milvus-haystack

from milvus_haystack import MilvusDocumentStore
from milvus_haystack.milvus_embedding_retriever import MilvusEmbeddingRetriever

document_store = MilvusDocumentStore(connection_args={"uri": "./milvus.db"}, drop_old=True,)
retriever = MilvusEmbeddingRetriever(document_store=document_store, top_k=3)

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 your own dataset to customize your RAG chatbot.

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

indexing_pipeline = Pipeline()
indexing_pipeline.add_component("converter", MarkdownToDocument())
indexing_pipeline.add_component("splitter", DocumentSplitter(split_by="sentence", split_length=2))
indexing_pipeline.add_component("embedder", document_embedder)
indexing_pipeline.add_component("writer", DocumentWriter(document_store))
indexing_pipeline.connect("converter", "splitter")
indexing_pipeline.connect("splitter", "embedder")
indexing_pipeline.connect("embedder", "writer")
indexing_pipeline.run({"converter": {"sources": file_paths}})

# print("Number of documents:", document_store.count_documents())
  
question = "What is Milvus?"  # You can replace it with your own question.

retrieval_pipeline = Pipeline()
retrieval_pipeline.add_component("embedder", text_embedder)
retrieval_pipeline.add_component("retriever", retriever)
retrieval_pipeline.connect("embedder", "retriever")

retrieval_results = retrieval_pipeline.run({"embedder": {"text": question}})

# for doc in retrieval_results["retriever"]["documents"]:
#     print(doc.content)
#     print("-" * 10)

from haystack.utils import Secret
from haystack.components.builders import PromptBuilder

retriever = MilvusEmbeddingRetriever(document_store=document_store, top_k=3) 
 text_embedder = SentenceTransformersTextEmbedder(model="BAAI/bge-large-en-v1.5")
text_embedder.warm_up()

prompt_template = """Answer the following query based on the provided context. If the context does
                     not include an answer, reply with 'I don't know'.\n
                     Query: {{query}}
                     Documents:
                     {% for doc in documents %}
                        {{ doc.content }}
                     {% endfor %}
                     Answer: 
                  """

rag_pipeline = Pipeline()
rag_pipeline.add_component("text_embedder", text_embedder)
rag_pipeline.add_component("retriever", retriever)
rag_pipeline.add_component("prompt_builder", PromptBuilder(template=prompt_template))
rag_pipeline.add_component("generator", generator)
rag_pipeline.connect("text_embedder.embedding", "retriever.query_embedding")
rag_pipeline.connect("retriever.documents", "prompt_builder.documents")
rag_pipeline.connect("prompt_builder", "generator")

results = rag_pipeline.run({"text_embedder": {"text": question}, "prompt_builder": {"query": question},})
print('RAG answer:\n', results["generator"]["replies"][0])

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.

Haystack optimization tips

To optimize Haystack in a RAG setup, ensure you use an efficient retriever like FAISS or Milvus for scalable and fast similarity searches. Fine-tune your document store settings, such as indexing strategies and storage backends, to balance speed and accuracy. Use batch processing for embedding generation to reduce latency and optimize API calls. Leverage Haystack's pipeline caching to avoid redundant computations, especially for frequently queried documents. Tune your reader model by selecting a lightweight yet accurate transformer-based model like DistilBERT to speed up response times. Implement query rewriting or filtering techniques to enhance retrieval quality, ensuring the most relevant documents are retrieved for generation. Finally, monitor system performance with Haystack’s built-in evaluation tools to iteratively refine your setup based on real-world query performance.

Milvus optimization tips

Milvus serves as a highly efficient vector database, critical for retrieval tasks in a RAG system. To optimize its performance, ensure that indexes are properly built to balance speed and accuracy; consider utilizing HNSW (Hierarchical Navigable Small World) for efficient nearest neighbor search where response time is crucial. Partitioning data based on usage patterns can enhance query performance and reduce load times, enabling better scalability. Regularly monitor and adjust cache settings based on query frequency to avoid latency during data retrieval. Employ batch processing for vector insertions, which can minimize database lock contention and enhance overall throughput. Additionally, fine-tune the model parameters by experimenting with the dimensionality of the vectors; higher dimensions can improve retrieval accuracy but may increase search time, necessitating a balance tailored to your specific use case and hardware infrastructure.

Anthropic Claude 3 Sonnet optimization tips

To optimize Claude 3 Sonnet in RAG workflows, refine retrieval chunk sizes to balance context relevance and token efficiency—aim for 500-800 token chunks with 15% overlap. Use structured prompts with XML tags or markdown to separate instructions from retrieved content, explicitly directing Claude to ground responses in provided sources. Lower temperature (0.2-0.4) improves factual consistency, while adding validation steps like “Verify this answer is fully supported by the context” reduces hallucinations. Prioritize system prompts to define response formats and enforce source citation. Test top-p (0.7-0.9) and max tokens to control output breadth without truncation.

BAAI bge-large-en-v1.5 optimization tips

To optimize BAAI bge-large-en-v1.5 in RAG, ensure input text is clean and preprocessed (remove noise, truncate to 512 tokens). Use the model’s instruction prefix ("Represent this sentence for retrieval: ") for query embeddings to align with its training. Batch embedding generation for efficiency, and normalize outputs before similarity comparisons. Fine-tune on domain-specific data if retrieval accuracy lags. Use FAISS or HNSW for fast vector search, and quantize embeddings to reduce memory. Regularly evaluate recall@k to balance speed and relevance. Leverage GPU acceleration and optimize temperature/sampling for generation coherence.

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?

By diving into this tutorial, you’ve unlocked the power of combining cutting-edge tools to build a fully functional RAG system from scratch! You learned how Haystack acts as the glue, seamlessly orchestrating your pipeline by connecting a Milvus vector database for lightning-fast similarity searches, Anthropic Claude 3 Sonnet for generating human-like, context-aware responses, and the BAAI bge-large-en-v1.5 embedding model to transform text into rich numerical representations. Together, these tools form a dynamic workflow: the embedding model encodes your data into vectors, Milvus stores and retrieves them efficiently, and Claude 3 synthesizes the retrieved knowledge into coherent, insightful answers. You saw firsthand how each component shines—whether it’s Claude’s knack for nuanced language, BAAI’s robust embeddings, Milvus’s scalability, or Haystack’s flexibility in tying it all together.

But this isn’t just about the basics—you also picked up pro tips for optimizing your RAG system, like fine-tuning chunk sizes for better retrieval or balancing cost and performance with Claude 3’s API. And let’s not forget the free RAG cost calculator you explored, a game-changer for budgeting your projects without surprises! Now that you’ve seen how these pieces fit into a cohesive, intelligent system, the real magic begins. Imagine building custom chatbots, research assistants, or domain-specific knowledge hubs—your creativity is the limit. So fire up your IDE, experiment with tweaks, and let your ideas take flight. The future of AI-powered applications is yours to shape, one RAG pipeline at a time. Let’s build something amazing! 🚀

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! 💖