Unlike most solutions that only use codebases, Kodesage goes beyond and integrates with various data sources to capture the collective knowledge of software projects. Key elements of the solution include an ingestion service to process and analyze the this knowledge, including existing documentation, source code, existing issue tickets, database schemas and tests.

The Kodesage Engine builds a dynamic knowledge base, capturing and generating new information using state-of-the-art techniques. It includes a knowledge graph that maps complex relationships between code, business domains, APIs, and databases, enhancing code comprehension and navigation. Kodesage Auto-sync keeps documentation current and synced with your code.

Kodesage also offers customizable enterprise-specific prompts and Retrieval Augmented Generation (RAG) to ensure the most relevant, up-to-date responses from Large Language Models (LLMs). Key features include a service to process and analyze source code, creating a searchable vector database and knowledge graph. This supports semantic code search, automated document generation, and issue ticketing integration, all powered by dedicated LLM instances for scalability and reliability.

Customers can access the knowledge generated in the Engine via easy-to-use Kodesage apps. These options include automatically generated documentation, a natural language interface for interaction, and issue guide integration with tools like Jira.

With “Ask Kodesage,” developers can leverage a chat feature to ask natural language questions about any technical aspects, simulating interaction with an experienced tech lead. The issue guide integration with tools like Jira suggests code modifications and implementation plans, streamlining the development process. Additionally, Kodesage offers automatic documentation generation and keeping it up-to-date.