DGi

DGi is a data platform designed specifically for developers who need to securely and accurately query structured data through natural language. It functions as an AI-powered data agent, allowing users to integrate its capabilities directly into their own applications via a simple REST API. The platform processes user questions about data stored in databases, data lakes, or other sources, and returns precise answers without requiring the developer to write complex SQL queries or manage query logic manually. The core workflow involves two primary components: session management and task execution. Developers first create a new session using a POST request to the `/session/new` endpoint. Within that session, tasks are created by submitting a question or query to the `/task/new` endpoint. Each task is interpreted by DGi’s underlying AI model, which analyzes the semantics of the question, maps it to the relevant data schema, and generates an accurate query to retrieve the answer. The response is then returned in a structured format, which the developer can present to end users within their own interface. DGi supports both internal and external client scenarios. For example, a company building a customer support dashboard can embed DGi so that support agents can ask questions like “Show me all open tickets from last week with priority high” and receive an instant answer without needing to know the underlying database structure. Similarly, external clients of a SaaS platform can use a branded chat interface powered by DGi to ask questions about their own data, such as sales figures or user activity logs, all while keeping the data within the security boundary of the developer’s infrastructure. The platform emphasizes reliability and accuracy. It uses a multi-layered validation