🔒 Privacy-First Philosophy
Your code stays on your machine. Code-y is fundamentally designed with your privacy and data security as its top priority. We understand the sensitivity of your codebase, and our architecture reflects this.
Core Tenets of Code-y's Privacy Model:
- 100% Local Processing by Default: All core analysis, including AST parsing and relationship mapping, happens entirely on your local workstation.
- Local AI with Ollama: When using the recommended Ollama integration, all AI model processing (for descriptions, chat, semantic search) occurs locally. Your code snippets are not sent to any external servers.
- Local Vector Embeddings: Code embeddings for semantic search are generated and stored in a local FAISS database within your project's `documentation` folder.
- No Unsolicited Data Transmission: Code-y does not transmit your code or any derived data to external services unless you explicitly configure and use the OpenAI integration with your own API key.
- Complete Offline Capability: Once Ollama and your desired local models are set up, Code-y can operate fully offline, providing all its analytical and AI-powered features without an internet connection.
- Full Data Control: You retain complete ownership and control over all generated documentation and analysis data, as it's stored locally on your file system.
- Optional Cloud Features: Our upcoming premium cloud services (like enhanced AI compute and HIPAA-compliant storage) are strictly opt-in. The core local-first functionality will always remain private and under your control.
When is Data Sent Externally?
The only instance where data might leave your local machine is if you:
- Explicitly configure Code-y to use the OpenAI API by providing your own API key. In this case, relevant code snippets are sent to OpenAI for processing according to their terms of service.
- Opt-in to use future Code-y Cloud services. Data handling for these services will be clearly outlined in their respective terms and privacy policies.
Code-y is committed to transparency. We believe you should always know where your data is and how it's being processed.