EmbeddingFramework

Modular • Extensible • Production-Ready
A Python framework for embeddings, vector databases, and cloud storage providers.

---

📚 Documentation

 

---

A modular, extensible, and production-ready Python framework for working with embeddings, vector databases, and cloud storage providers.
Designed for AI, NLP, and semantic search applications, EmbeddingFramework provides a unified API to process, store, and query embeddings across multiple backends.

---

✨ Features

🔹 Multi-Vector Database Support

• ChromaDB – Local and persistent vector storage.
• Milvus – High-performance distributed vector database.
• Pinecone – Fully managed vector database service.
• Weaviate – Open-source vector search engine.

🔹 Cloud Storage Integrations

• AWS S3 – Store and retrieve embeddings or documents.
• Google Cloud Storage (GCS) – Scalable object storage.
• Azure Blob Storage – Enterprise-grade cloud storage.

🔹 Embedding Providers

• OpenAI Embeddings – State-of-the-art embedding generation.
• Easily extendable to other providers.

🔹 File Processing & Preprocessing

• Automatic file type detection.
• Text extraction from multiple formats including .txt, .pdf, .docx, .csv, .xls, .xlsx.
• Preprocessing utilities for cleaning and normalizing text.
• Intelligent text splitting for optimal embedding performance.
• Large dataset handling for Excel files with efficient chunking to preserve embedding context.

🔹 Utilities

• Retry logic for robust API calls.
• File utilities for safe and efficient I/O.
• Modular architecture for easy extension.

---

📦 Installation & Setup

# Basic installation
pip install embeddingframework

# With development dependencies
pip install embeddingframework[dev]

---

⚡ Quick Start Example

from embeddingframework.adapters.openai_embedding_adapter import OpenAIEmbeddingAdapter
from embeddingframework.adapters.vector_dbs import ChromaDBAdapter

# Initialize embedding provider
embedding_provider = OpenAIEmbeddingAdapter(api_key="YOUR_OPENAI_API_KEY")

# Initialize vector database
vector_db = ChromaDBAdapter(persist_directory="./chroma_store")

# Generate embeddings
embeddings = embedding_provider.embed_texts(["Hello world", "EmbeddingFramework is awesome!"])

# Store embeddings
vector_db.add_texts(["Hello world", "EmbeddingFramework is awesome!"], embeddings)

---

📂 Project Structure

embeddingframework/
│
├── adapters/                # Vector DB & storage adapters
│   ├── base.py
│   ├── chromadb_adapter.py
│   ├── milvus_adapter.py
│   ├── pinecone_adapter.py
│   ├── weaviate_adapter.py
│   ├── storage/             # Cloud storage adapters
│
├── processors/              # File processing logic
├── utils/                    # Helper utilities
└── tests/                    # Test suite

---

🧪 Testing

pytest --maxfail=1 --disable-warnings -q

With coverage:

pytest --cov=embeddingframework --cov-report=term-missing

---

🔄 CI/CD

This project includes a GitHub Actions workflow (.github/workflows/python-package.yml) for:

• Automated testing with coverage.
• Version bumping & changelog generation.
• PyPI publishing.
• GitHub release creation.

---

📜 License

This project is licensed under the MIT License – see the LICENSE file for details.

---

🤝 Contributing

Contributions, issues, and feature requests are welcome!
Feel free to check the issues page.

1. Fork the repository.
2. Create a new branch (feature/my-feature).
3. Commit your changes.
4. Push to your branch.
5. Open a Pull Request.

---

🌟 Why EmbeddingFramework?

• Unified API – Work with multiple vector DBs and storage providers seamlessly.
• Extensible – Add new adapters with minimal effort.
• Production-Ready – Built with scalability and reliability in mind.
• Developer-Friendly – Clean, modular, and well-documented codebase.

---

📖 Full Documentation Overview

Below is a comprehensive, end-to-end guide covering all features, usage patterns, and advanced configurations of EmbeddingFramework.

1️⃣ Introduction

EmbeddingFramework is designed to simplify the integration of embeddings, vector databases, and cloud storage into AI-powered applications. It provides:

• A unified API for multiple backends.
• Extensible architecture for adding new providers.
• Production-ready reliability with retries, error handling, and modular design.

---

2️⃣ Installation

pip install embeddingframework
pip install embeddingframework[dev]  # For development

---

3️⃣ Supported Vector Databases

Database	Type	Key Features
ChromaDB	Local	Persistent storage, lightweight
Milvus	Distributed	High-performance, scalable
Pinecone	Managed	Fully hosted, easy to scale
Weaviate	Open-source	Semantic search, hybrid queries

---

4️⃣ Cloud Storage Integrations

EmbeddingFramework supports:

• AWS S3
• Google Cloud Storage
• Azure Blob Storage

Example:

from embeddingframework.adapters.storage.s3_storage_adapter import S3StorageAdapter
storage = S3StorageAdapter(bucket_name="my-bucket")
storage.upload_file("local.txt", "remote.txt")

---

5️⃣ Embedding Providers

Currently supported:

• OpenAI Embeddings
• Easily extendable to HuggingFace, Cohere, etc.

Example:

from embeddingframework.adapters.openai_embedding_adapter import OpenAIEmbeddingAdapter
provider = OpenAIEmbeddingAdapter(api_key="YOUR_KEY")
embeddings = provider.embed_texts(["Hello", "World"])

---

6️⃣ File Processing

EmbeddingFramework provides a robust and extensible file processing pipeline that can handle a wide variety of file formats and sizes. This includes:

• Automatic File Type Detection – The framework automatically determines the file type and routes it to the appropriate parser.
• Text Extraction – Supports extracting text from:
  • .txt – Plain text files
  • .pdf – PDF documents
  • .docx – Microsoft Word documents
  • .csv – Comma-separated values
  • .xls / .xlsx – Microsoft Excel spreadsheets (including multi-sheet workbooks)
• Preprocessing Utilities – Cleans and normalizes extracted text for better embedding quality (e.g., removing stopwords, normalizing whitespace).
• Intelligent Text Splitting – Splits large documents into smaller, context-friendly chunks for optimal embedding performance.
• Large Dataset Handling for Excel – Efficiently processes large Excel files by:
  • Reading all sheets in the workbook.
  • Converting each row into a string representation.
  • Chunking rows into manageable segments to avoid exceeding embedding context limits.
  • Applying quality filters to remove empty or low-value chunks.

This design ensures that even massive datasets can be processed without memory overload or loss of semantic context.

Example:

from embeddingframework.processors.file_processor import FileProcessor

processor = FileProcessor()

# Process a PDF
pdf_text = processor.process_file("document.pdf")

# Process a large Excel file with multiple sheets
excel_text = processor.process_file("large_dataset.xlsx")

# Process a CSV file
csv_text = processor.process_file("data.csv")

# Process a DOCX file
docx_text = processor.process_file("report.docx")

Advanced Usage:

# Asynchronous processing with custom chunk sizes and quality filters
import asyncio

async def process_files():
    await processor.process_file_async(
        "large_dataset.xlsx",
        chunk_size=2000,
        text_chunk_size=1000,
        merge_target_size=3000,
        parallel=True,
        min_quality_length=50
    )

asyncio.run(process_files())

---

7️⃣ Utilities

• Retry logic
• File utilities
• Preprocessing helpers

---

8️⃣ CLI Usage

EmbeddingFramework includes a CLI:

embeddingframework --help

---

9️⃣ Advanced Configurations

• Custom vector DB adapters
• Custom embedding providers
• Batch processing
• Async support

---

🔟 End-to-End Example

from embeddingframework.adapters.openai_embedding_adapter import OpenAIEmbeddingAdapter
from embeddingframework.adapters.vector_dbs import ChromaDBAdapter

provider = OpenAIEmbeddingAdapter(api_key="KEY")
db = ChromaDBAdapter(persist_directory="./store")

texts = ["AI is amazing", "EmbeddingFramework is powerful"]
embeddings = provider.embed_texts(texts)
db.add_texts(texts, embeddings)

---

📊 Feature Matrix

Feature	Supported
Multi-DB Support	✅
Cloud Storage	✅
File Processing	✅
Retry Logic	✅
CLI	✅
Async	✅

---

📚 Learn More

For the full documentation, visit:
👉 EmbeddingFramework Docs