Welcome to PureCPP, where efficiency and optimization are the foundation of everything we build. Here, we value clean, fast, and powerful code.
Want to contribute? Whether you're creating new integrations, improving performance, or expanding features, every line of code matters. Together, we’ll take high-performance computing to the next level.
💡 Ready to code without limits? Let’s get to work! 💡
Ready to jump in? Follow this quick setup guide to get started smoothly:
-
Fork the repo and clone your fork.
-
Navigate to the project folder:
cd purecpp -
Make sure you have the following packages installed:
- GCC/G++ 13.1
- CMake 3.22+
- Conan 2
- Rust
- Python 3.8+
-
Install the required dependencies:`: Depending on your system, you may need:
sudo apt update && sudo apt install -y gcc-13 g++-13 cmake conan rustc cargo -
Install development conan dependencies::
!pip install conan==2.*- Run the tests to ensure everything is working::
./tests/run_tests
All set! Now it's time to build something powerful. If you need more details, check out the Development Guidelines.
Join our community Discord to ask questions, get support, and collaborate with fellow contributors and users.
There are many ways to contribute to PureCPP—whether you're a C++ expert or just starting out with high-performance computing. Here, we focus on performance, efficiency, and scalability. Your contributions are always welcome!
Help us improve PureCPP by contributing to our core modules and making the framework even more optimized.
- New Integrations (e.g., support for new compilers, optimized bindings, high-performance libraries)
- Memory Management, Parallelism (Threads and CUDA), Matrix and Tensor Operations
- Advanced Chunking Techniques to optimize processing
- Efficient Metadata Extraction and Management
- Optimized Dataloaders for different file types and databases
- Efficient indexing and retrieval
- Smart loading strategies to optimize search performance
- Implementation and optimization of high-performance vector databases
- Integration of LLMs and embedding models for semantic search
- Support for quantization, fine-tuning, and CUDA optimizations
Found something that could be optimized? Code improvements are always welcome! Check out the GitHub Labels
If you’ve used PureCPP in an innovative way, share your examples and contribute to the community.
Got a different idea? We’re open to tests and new approaches—experiment and submit a PR!
We are always evolving! Here are the next steps to make our pipelines even more efficient and powerful:
✅ Add local Vector Databases to enhance semantic search performance
✅ Integrate local LLMs and create connectors for inference frameworks
🛠️ Optimize data extraction for greater efficiency
📌 Add Schema to better structure data
📌 Expand the variety of models in our components
🔄 Enhance chunking techniques for smarter processing
📈 Improve embeddings for more precise vector representations
🗂️ Refine metadata extraction for better contextualization
💡 Got an idea? Your contribution is more than welcome! Join us and help take this project even further. 🚀
- Fork the repository on GitHub.
- Clone your fork to your local machine.
git clone https://github.com/pureai-ecosystem/purecpp.git
- Create a branch for your work.
git checkout -b your-feature-branch
- Set up your environment
- Work on your feature or bugfix, ensuring you have unit tests covering your code.
- Commit your changes, then push them to your fork.
git push origin your-feature-branch
- Open a pull request on GitHub.
Obrigado!
Big thanks for being part of PureCPP—where every bit counts, and every byte makes a difference! 🚀
Whether you're optimizing loops, fine-tuning embeddings, or pushing parallel processing to the limit, your contributions fuel the engine of high-performance computing.
We’re not just writing code—we’re compiling the future. 🔥
Keep coding at full speed! 🏎️💻