Build RAG Chatbot with Haystack, Pgvector, Anthropic Claude 3.5 Haiku, and Ollama bge-m3

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.
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.)
Anthropic Claude 3.5 Haiku: A lightweight, high-speed AI model optimized for rapid, cost-efficient processing of complex queries. It excels in real-time applications, offering strong performance in text analysis, summarization, and multilingual tasks. Ideal for scalable enterprise solutions, dynamic customer interactions, and scenarios requiring low-latency responses without compromising accuracy.
Ollama BGE-M3: A multilingual embedding model optimized for semantic understanding, retrieval, and clustering. Strengths include high accuracy across 100+ languages, robust performance in dense retrieval tasks, and scalability. Ideal for enterprise search systems, cross-lingual applications, and AI-driven knowledge management requiring precise semantic analysis.

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 3.5 Haiku

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-haiku-20240307")

Step 3: Install and Set Up Ollama bge-m3

pip install ollama-haystack

Make sure that you have a running Ollama model (either through a docker container, or locally hosted). No other configuration is necessary as Ollama has the embedding API built in.

from haystack import Document
from haystack_integrations.components.embedders.ollama import OllamaDocumentEmbedder
from haystack_integrations.components.embedders.ollama import OllamaTextEmbedder

text_embedder = OllamaTextEmbedder(model="bge-m3")
document_embedder = OllamaDocumentEmbedder(model="bge-m3")

Step 4: Install and Set Up Pgvector

To quickly set up a PostgreSQL database with pgvector, you can use Docker:

docker run -d -p 5432:5432 -e POSTGRES_USER=postgres -e POSTGRES_PASSWORD=postgres -e POSTGRES_DB=postgres ankane/pgvector

To use pgvector with Haystack, install the pgvector-haystack integration:

pip install pgvector-haystack

import os
from haystack_integrations.document_stores.pgvector import PgvectorDocumentStore
from haystack_integrations.components.retrievers.pgvector import PgvectorEmbeddingRetriever

os.environ["PG_CONN_STR"] = "postgresql://postgres:postgres@localhost:5432/postgres"

document_store = PgvectorDocumentStore()
retriever = PgvectorEmbeddingRetriever(document_store=document_store)

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 = PgvectorEmbeddingRetriever(document_store=document_store) 
 text_embedder = OllamaTextEmbedder(model="bge-m3")

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.

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.

Anthropic Claude 3.5 Haiku optimization tips

Optimize context relevance by implementing semantic chunking with 512-1024 token segments and metadata filtering to reduce noise. Use structured data formatting (JSON/XML tables) for retrieved content to enhance Claude's parsing efficiency. Implement query expansion with synonyms and domain-specific terms to improve retrieval alignment. Fine-tune temperature settings (0.2-0.4 range) for factual consistency. Cache frequent query embeddings and employ prompt compression techniques like entity summarization. Monitor token usage through Claude's API metrics to balance context depth with cost efficiency, prioritizing critical information in the prompt's beginning for better attention focus.

Ollama bge-m3 optimization tips

To optimize Ollama bge-m3 in a RAG setup, ensure input text is preprocessed (lowercasing, removing noise) and split into semantically coherent chunks (300-500 tokens) to balance retrieval accuracy and computational load. Use dynamic pooling to prioritize key phrases in embeddings. Fine-tune the model on domain-specific data with contrastive learning to enhance relevance. Adjust temperature and top-k sampling for controlled generation. Leverage batch inference for parallel processing and enable hardware acceleration (e.g., CUDA) for faster embeddings. Regularly validate retrieval performance with benchmark datasets to refine thresholds and ranking strategies.

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 create a robust RAG system! You’ve seen how Haystack acts as the backbone, orchestrating workflows and connecting components seamlessly. With Pgvector as your vector database, you learned to store and retrieve embeddings efficiently, ensuring lightning-fast similarity searches that ground your AI’s responses in real data. The magic truly comes alive with Anthropic Claude 3.5 Haiku, which leverages its speed and nuance to generate human-like answers, while Ollama’s bge-m3 embedding model transforms text into rich numerical representations—capturing meaning and context like never before. Together, these pieces form a dynamic pipeline that pulls from external knowledge and crafts responses that feel both intelligent and relevant. Plus, you’ve picked up pro tips like optimizing chunking strategies and tweaking retrieval parameters to balance speed and accuracy, and that free RAG cost calculator? A game-changer for keeping your projects budget-friendly without sacrificing performance!

Now that you’ve seen how these tools harmonize, imagine what you can build next! Whether it’s a customer support bot, a research assistant, or a creative brainstorming partner, you’ve got the skills to bring ideas to life. Remember, every tweak and experiment pushes the boundaries of what your RAG system can do. So fire up your code editor, play with those optimizations, and let your curiosity lead the way. The future of intelligent applications is in your hands—go make it awesome. Happy building! 🚀

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! 💖