Marie Robin
The danger isn't the tool. It's the habit.
Agencies have years of built-up instinct and client knowledge. That intelligence is the only thing AI can't replicate. But when teams accept "good enough" outputs without surfacing their prior research and judgment, they lose their edge.
The real risk isn't replacement. It's bad habits forming faster than good ones. I know where the intelligence lives inside an agency—and I design the systems to make it stick.
"How to save time with AI without losing focus? 🚀 Here are the best quick wins for marketing teams. The difference between an AI agent and simple automation is..."
Access Post ↗"With AI, you can arrive at a client meeting with 30 visual directions or 30 different angles... It completely changes the dynamic. As long as you have the right people..."
Access Post ↗"Most teams trained on AI in 2023, bought every subscription, and now can't tell which tools to keep. ChatGPT is no longer the default. In 2026, results come from systems, not isolated prompts."
Read on Substack ↗Shaped by experiences
Revisit some of my memories to get a sense of what made me me. A reminder that we're all the sum of experiments and sensations that AI will never understand.
AI
Training
With a background in digital marketing and social media, I understand the tools and use cases that actually matter. I train agency teams and marketing departments on the AI workflows that fit their reality, not generic demos.
AI System
Design
Designing production-ready AI workflows for agencies and marketing teams: content automation, AI agents for reporting and strategic watch, assisted slides production, and decision support. Every system built to run with your team, not without it.
Visit Fleet Forward ↗Keynote
Speaking
Addressing decision-makers on human-in-the-loop AI adoption, moving beyond simple tool demos to discuss true paradigm shifts.
System Queries
What does an AI transformation consultant for agencies actually do?
+I diagnose where AI creates real value in your agency's operations, design custom workflows that fit your team's reality, and implement systems that stick. The work spans strategy alignment with leadership, hands-on training for creative and commercial teams, and building automated pipelines for recurring tasks. The goal is operational autonomy, not tool adoption.
How is AI reshaping the creative agency business model?
+Clients now expect faster delivery at lower cost. 58% of advertisers plan to reduce agency fees because of AI. Agencies that only sell production are becoming commodities. The agencies winning in 2026 sell strategic intelligence, proprietary systems, and the ability to orchestrate AI-augmented creative workflows their clients can't build alone.
What's the difference between AI training and AI transformation?
+Training teaches tools. Transformation changes how work gets done. Most agencies have done a one-day ChatGPT workshop — six weeks later, everyone's back to old habits. Transformation means diagnosing workflows, building custom AI systems, training on real use cases, and following up until adoption is measurable. Training is a day. Transformation is a 2-month engagement with lasting impact.
What is an Intelligence Factory?
+An Intelligence Factory is a system that digitizes your agency's collective intelligence — client knowledge, creative processes, strategic frameworks — and makes it accessible to your team and their AI assistants. It combines automated performance monitoring, a knowledge wiki, and calibrated content generation relying on agency expertise. The result: your team produces faster without losing quality or brand consistency.
What is a Brand OS?
+A Brand OS is the operating infrastructure that lets marketing teams produce content quickly, accurately, and without waiting for an agency or internal expert for every deliverable. It centralizes brand memory, tone guidelines, and approved assets into a system that AI can access in real time. Teams become autonomous producers instead of permanent requesters.
How long does an AI transformation take for a creative agency?
+A diagnostic takes 2 weeks. A first operational pilot runs 4 to 8 weeks. Full transformation with measurable adoption typically takes 3 to 6 months depending on team size and complexity. The approach is incremental: we start with one use case that delivers a visible before/after, then scale what works.