Daniel Corydon-Petersen

Daniel Corydon-Petersen is an AI consultant at Gladsaxe Municipality and works on creating value through AI across the organisation. Before joining Gladsaxe, he worked as a data scientist in ATP's AI department, where he built AI solutions. His technical background shapes his approach: he connects practical experimentation with critical questions about implementation, responsibility and long-term maintenance, and he is interested in how AI tools can fit into employees' actual work processes and be used thoughtfully.

In this work, he draws on hybrid intelligence as formulated by the Center for Hybrid Intelligence at Aarhus University: AI and people participate in the same workflow with a deliberate division of labour, where AI handles what it does well, such as finding patterns, reviewing information and producing first drafts, while employees retain judgement, weigh the considerations and remain responsible. The question is not how much can be automated, but how AI and employees together can solve tasks better than either could alone, in a way where the human learns more than the machine.

With his technical background, he also actively explores the value that vibe coding can create in the municipality and how it should be supported within the organisation. He sees an opportunity to develop close to users and address smaller local needs, and therefore focuses on what this requires: skills in hosting, maintenance, security updates, training, etc.

Finally, he works on questions of digital sovereignty: where the municipality obtains its LLM inference, what the choice between open-source and closed-source models and systems means, and how dependence on particular models and suppliers can be managed so that the municipality retains control over its data and choices in the longer term.

Selected experience

• AI consultant at Gladsaxe Municipality and former data scientist in ATP's AI department.
• Hybrid intelligence with a deliberate division of labour between AI and employees.
• Vibe coding close to users, with attention to hosting, maintenance, security updates and training.
• Digital sovereignty, LLM inference and dependence on models and suppliers.