Every time a new AI capability launches, someone writes the piece about how it will replace a professional class. Lawyers. Radiologists. Now: agile coaches. The piece is always wrong in the same way — it confuses automation of tasks with replacement of judgement.
What AI actually does in a coaching context
Albert AI — Performalise's always-on AI coach — monitors team health signals, Agile Event quality, retro action completion, and sprint pattern data continuously. It surfaces anomalies before coaches would typically notice them, prepares Agile Event briefs that take account of the last three sprints of context, and prompts teams with questions calibrated to their specific situation.
What it doesn't do is sit with a team that's quietly falling apart because two senior engineers haven't spoken in three weeks. It doesn't read the room. It doesn't adapt in real time to someone's tone of voice.
"The question is never AI or human. It's always: what does each do better, and how do you connect them?"McKinsey Global Institute · Future of Work
The maths of scale
The average senior agile coach supports six to ten teams. At forty hours a week, that's roughly four to six hours per team — enough for Agile Event facilitation and one meaningful conversation. Not enough for continuous health monitoring, data analysis, pattern recognition across teams, and personalised coaching nudges.
Albert AI handles the monitoring layer continuously, at zero marginal cost per team. This gives coaches back the hours they were spending on preparation and status assessment, and redirects them toward the work only humans can do: building trust, navigating conflict, and developing team capability over time.
The result isn't replacement. It's a coach who can meaningfully support twelve teams instead of six — with better outcomes for all twelve.