Marcelo Fernandes
Marcelo Fernandes — Director of Data & AI Strategy | Dynamics 365, Fabric & Copilot
São Paulo, Brazil
Actions
Marcelo Fernandes is a data and AI superhero disguised as a business leader. By day, he helps organizations turn chaos into clarity with Microsoft Fabric, Power Platform, Dynamics 365, and Copilot. By night, he builds AI agents, governs data empires, and fights bad architecture wherever it hides. With 20+ years in tech and a passion for turning innovation into impact, Marcelo’s mission is simple: make AI work for humans, not the other way around.
Area of Expertise
Why Forecasts Fail—and How to Build AI Models Leaders Actually Use
Forecasting is often the first place organizations apply AI — and also where adoption fails the fastest.
In this session, we explore why many statistical and machine learning forecasts never make it into real operational decision-making, even when accuracy looks good on paper. Using Dynamics 365, Power BI, and Microsoft Fabric as reference platforms, we focus on how to build predictive models that business leaders actually trust and act on.
Topics include:
-) The core reasons AI forecasts struggle in live supply chain operations
-) Practical techniques to quantify, communicate, and visualize uncertainty
-) Ways to explain model outputs so planners and leaders feel confident adjusting decisions
-) Operational feedback loops that continuously improve forecasts over time
This session goes beyond models and metrics to address the real challenge: behavioral adoption. Because forecasting only creates value when people believe it — and change plans because of it.
Ideal for supply chain leaders, architects, analysts, and consultants looking to move AI forecasting from “interesting dashboards” to trusted operational tools.
Copilot in the Real World: 10 AI Use Cases You Can Deploy in 60 Days
Copilot is no longer a novelty — but many organizations are still stuck in demo mode.
In this session, we walk through 10 practical Copilot and AI use cases built on Dynamics 365 that can realistically be implemented within 30, 60, or 90 days, depending on organizational maturity.
Rather than generic examples, this talk focuses on real operational scenarios, including:
-) Intelligent order prioritization in Dynamics 365 F&O
-) AI-assisted explanations for production deviations
-) Copilot support for GRNI discrepancy resolution
-) Forecast augmentation with human-in-the-loop validation
-) For each use case, we outline:
-) The business problem being solved
-) The architectural pattern (Dynamics, Copilot, Azure AI, Fabric)
-) Success metrics and expected impact
-) A practical implementation roadmap
The goal is simple: give you a menu of bite-sized, high-impact Copilot initiatives you can start building next week — without waiting for a multi-year AI transformation program.
Ideal for architects, consultants, and business leaders who want Copilot to deliver measurable value fast.
AI That Decides: Closing the Data-to-Action Gap in Dynamics 365
Most organizations don’t suffer from a lack of data — they struggle to turn insights into consistent, repeatable business decisions.
In this session, we explore how manufacturers and distributors can operationalize AI across their Dynamics 365 landscape, transforming analytics into real actions such as reorder triggers, exception handling, and predictive maintenance workflows.
You’ll see practical AI patterns and architectural building blocks that connect Dynamics 365 with Microsoft Fabric and Azure AI to deliver:
-) Automated alerts and recommendations
-) Human-in-the-loop decision workflows
-) Actionable insights embedded directly into operational processes
Rather than focusing on hype or dashboards alone, this session emphasizes decisioning: how AI supports better, faster decisions while keeping humans in control.
We’ll also discuss how to measure impact through KPIs such as lead time reduction, OEE improvement, and forecast accuracy — helping teams move from experimentation to measurable ROI.
Ideal for solution architects, functional consultants, and leaders who want AI that actually changes how work gets done.
Please note that Sessionize is not responsible for the accuracy or validity of the data provided by speakers. If you suspect this profile to be fake or spam, please let us know.
Jump to top