News from Mastering Snowflake

We've just completed the first of four customised Snowflake training sessions with the Australian Football League (AFL) data team, held onsite at Marvel Stadium.

This session focused on Snowflake Notebooks: iterative analysis using markdown, Python, and SQL, plus building, scheduling, and collaborating through the Workspace feature. Rather than working through generic examples, the team trained directly on real AFL data, which made the concepts click faster and sparked genuinely insightful questions throughout the day.

Three more sessions are on the calendar, covering Streamlit, Cortex AI, and Snowflake CoWork.

It's a great reminder of the value of hands-on, customised education: teams walk away with genuine confidence, not just exposure.

The shift from pipeline coder to data product builder doesn't happen automatically. This 3-part video series, hosted by Adam for Matillion's Maia platform, gives data engineers a clear view of how agentic systems work, an honest framework to evaluate their own role, and the skills needed to stay relevant as the work changes.

Episode 1, "Your Role is Changing, and It's a Good Thing" (31 min), unpacks why data teams are hitting a sustainability wall and what an agentic system actually is, with an interactive assessment tool to help you gauge where your own role sits today.

Episode 2, "From Coder to Strategist" (23 min), lays out a practical framework for auditing your current workload and shifting toward AI-augmented, architecture-led work, paired with an interactive Manual Tax Audit tool to quantify how much time repetitive tasks are really costing you.

Episode 3, "Agentic AI in Action: Real World Transformations" (12 min), features an interview with Chris Mihalicz, CEO of Three Point Turn, on achieving five to ten times the output per head through automation, and closes with a 30-day implementation checklist.

Adam's new O'Reilly report, Beyond Pipelines: The New Data Operating Model for AI at Scale, is now available.

As enterprises move AI from pilots into production, fragmented data architectures create hidden costs. Disconnected systems, siloed teams, and reactive governance quietly degrade AI performance and erode agility and trust. The report makes the case that fixing this fragmentation is a strategic imperative, not a technical nice-to-have, for any organisation trying to scale AI responsibly.

It sets out what a unified, AI-ready data foundation actually looks like, and the organisational shift needed to support agentic AI: moving from siloed ownership to cross-functional teams, coordination models, and governance frameworks that evolve from compliance roadblocks into enablers of autonomy.

The central idea is reframing governance as trust infrastructure. Get that right, and it becomes the foundation organisations need to scale AI with confidence rather than the thing slowing them down.