The sales tool is a new development for the centralized management of energy-industry-related, metering-point-specific data and the seamless mapping of the sales and contract processes. As the system of record, it is designed with a strong focus on integration and exchanges data with numerous surrounding systems via a wide range of interfaces. Throughout the implementation, AI-assisted development using the GitHub Copilot CLI as an agentic assistant is consistently applied. The AI assistant is firmly integrated into the daily development process through project-wide guidelines and recurring shortcuts and, under human supervision, performs research, code generation, refactoring, testing, and reviews. This creates a continuous AI-assisted development workflow across multiple microservices and a web frontend. The goal of using AI is to deliver solutions faster, with higher quality, while optimizing costs.
Technically, the sales tool consists of a web frontend, a BFF gateway, and several microservices, each with its own relational database. The components are operated in a public cloud and connected to surrounding systems via a message bus, REST APIs, and OData. AI assistance is firmly integrated into the workflow: The command-line AI automatically loads an AI harness consisting of rules, templates, skills, and role-based agents and works with specialized sub-agents for research, implementation, and review. A file-based working memory, a worktree-per-branch model, and tool integrations via a Model Context Protocol make the AI sessions reproducible and consistent across multiple parallel topics. With this AI assistance, the entire process chain from specification, planning, implementation, and testing through deployment, error analysis, and issue resolution can be covered, accompanied by continuous optimization of the development process.
From a functional perspective, the sales tool covers the entire end-to-end process, from sales through quotation and contract creation to the end of the contract term, and serves as the system of record for energy-industry-related, metering-point-specific data. The focus is on the AI-assisted implementation of functional requirements: With AI support, testable user stories, feasibility analyses, technical proposals, in-depth research, plans, and executable code are created from workshops, specifications, and tickets – strictly specification-driven, without inventing requirements, and with a clear separation between facts, assumptions, and open questions. AI accelerates domain analysis, data modeling, and interface development, while functional decisions and quality assurance remain under human responsibility.
