Sourcegraph Cody

About Sourcegraph Cody

Sourcegraph was founded in 2013 with a focus on code search and intelligence—helping developers understand and navigate large, complex codebases. Cody is Sourcegraph's AI coding assistant, launched in 2023, and its key differentiator is the deep codebase context it can provide. Rather than relying only on files currently open in the editor, Cody uses Sourcegraph's code graph to retrieve relevant code snippets from across an entire codebase at query time, delivering more accurate and contextually appropriate AI suggestions for large enterprise repositories. Cody supports multiple underlying LLM providers—including Anthropic Claude, OpenAI GPT-4, and Google Gemini—and can be configured to use an organisation's preferred model or even a self-hosted LLM endpoint. This flexibility is particularly valuable for enterprises with data sovereignty requirements: the LLM processing can be directed to an approved provider or to an on-premise model deployment, while Sourcegraph handles the code retrieval and context assembly. For European enterprise engineering teams, Cody's enterprise architecture offers meaningful compliance controls. Self-hosted deployment of the Sourcegraph platform (on AWS, GCP, Azure, or bare metal) keeps all code and query data within the organisation's own infrastructure. Enterprise customers can bring their own LLM provider keys, ensuring that AI processing happens under their own contracts with their chosen model provider. SSO, admin controls, audit logging, and GDPR-compliant DPAs are available for enterprise accounts. Sourcegraph is US-incorporated but the enterprise self-hosted architecture enables European businesses to deploy Cody in a way that satisfies EU data residency and sovereignty requirements. This makes it one of the more genuinely enterprise-ready AI coding assistants for regulated European organisations.