AI and Productivity: Enedis trials the “Augmented Developer”

Paving the way for AI-driven development

Enedis is accelerating its digital transformation by testing AI-driven development tools. Early results show a clear impact on productivity, the quality of deliverables, and the professional growth of technical teams.

Moving towards standardised code

As part of its strategy to modernise information systems, Enedis is exploring the benefits of artificial intelligence within its technical departments. The company has launched a pilot project using GitHub Copilot, a development assistant that integrates directly into programming environments.

This intelligent assistant automatically generates lines of code, suggests fixes, detects inconsistencies, and recommends performance or style improvements. In practice, the tool analyses the context of a project and anticipates the next lines of code. It acts as a digital co-author for the programmer.

The trial involves several IT teams. Its goal is to measure the actual impact of AI on productivity, software quality, and developer satisfaction.

Promising and measurable results

Initial feedback is encouraging. Enedis is already observing productivity increases of up to 15% in development phases and 20% in testing. The AI assistant helps to quickly generate repetitive functions, fix syntax errors before they become blockers, and produce more consistent technical documentation.

For junior developers, the tool acts as an intelligent mentor by suggesting alternative approaches and explaining code snippets. This support aids continuous learning. Meanwhile, more experienced specialists can focus on higher-value tasks such as architectural design, software engineering, and performance optimisation.

Support ticket analysis has also improved. AI helps to identify recurring bug patterns and prioritise fixes. This reduces the average time taken to resolve issues.

Early estimates suggest that if these gains were applied across all teams, the impact would represent tens of thousands of hours saved per year. This would accelerate delivery cycles and reduce maintenance costs while improving software quality.

A structured and controlled approach

For Enedis, this initiative is part of a pragmatic approach to change: test, measure, and then scale. The company remains focused on data security, intellectual property, and regulatory compliance. These factors are vital in the energy sector.

Before any large-scale deployment, technical and organisational safeguards are established. These include defining specific use cases, testing AI behaviour in isolated environments, and supporting teams through awareness sessions. This rigorous governance allows the company to combine the power of AI tools with the reliability and traceability required of a major industrial operator.

A new generation of developers

Beyond productivity gains, the real focus is on people. The augmented developer does not replace a professional. Instead, the technology enhances their capability. AI becomes a cognitive assistant that handles repetitive tasks. This frees up time for reflection, architecture, and system design.

Developers are evolving into more analytical roles. These roles involve supervising AI models, evaluating automated suggestions, and performing final code validation. This hybrid skill set combines programming, data, and AI. It signals a lasting shift in software development careers.

Attracting attention across the industry

The Enedis project reflects a wider trend. According to a 2025 GitHub study, more than 60% of technology and industrial companies are already testing or deploying AI assisted development tools.

The most common benefits reported include improved code quality and a faster time to market. However, these initiatives also raise important questions. Organisations must consider how to verify the reliability of AI suggestions and how to maintain internal coding standards. This is where the role of specialised partners becomes essential.

The Infotel perspective

Infotel is already supporting several organisations as they adopt augmented developer tools. Our own pilots involving GitHub Copilot and automated testing assistants show performance gains of up to 20% in design and validation phases.

Our value lies in how we structure this usage. We focus on securing environments, protecting sensitive data, and defining relevant use cases. Most importantly, we train teams to benefit from AI while maintaining full control over code quality.

At Infotel, we turn experimentation into measurable benefits. We place every project into a framework of long-term performance and successful scaling.