Julian Wiffen

Meet Sam. A year ago she built data pipelines. Today she runs a data team she didn't hire. A load fails at 2am; AI Agents investigate and leave a fix for her approval by 9am. The migration nobody wanted to own moves through in the background, while Sam works the "needs attention" items. Once work arrives faster than one person can check it, the constraint stops being what agents can do and starts being how far you trust them. This is a day in the life of the agentic data team, and it changes what "data engineer," "manager," and "data leader" each mean. Join Matillion's Julian Wiffen and Cyril Sonnefraud to see how continuous context and autonomous execution turn every contributor into a team lead — and why that's the most important org change data leaders will navigate this decade.

The anomaly lived for three weeks between the QBR metric definition and pipeline logic. Nobody caught it until the number was in the board pack. The alternative is Maia: a system where context builds as your team works. The Context Engine captures institutional knowledge from Slack, analyst notes and BRDs, building context around metrics, rules, exceptions and naming conventions. Mission Control runs agents that monitor data movements against that context, spots mismatches and starts the investigation. We’ll show Maia catching a data movement, explaining what changed, then handing Snowflake CoWork a structured path for deeper exploration.

This session will explore the evolving role of data engineers. Data engineering is currently a bottleneck due to overwhelming requests and complex knowledge work. Maia acts as a "digital data engineer" or a "virtual data team" that amplifies productivity by 100x. It enables users, from skilled engineers to citizen data analysts, to author pipelines in natural business language. The session will demonstrate Maia's ability to accelerate mundane and advanced tasks,troubleshoot and debug pipelines in real-time, and generate high-quality, auditable pipelines using Matillion's proprietary, human-readable Data Pipeline Language (DPL), which avoids "spaghetti code" common with generic LLMs.

A paradigm shift is underway; the primary consumer of data is evolving from human analysts to AI agents. This presents a strategic challenge to every data leader: how do we architect an ecosystem that satisfies relentless, machine-scale demand for governed data without overwhelming our most valuable human experts? A chaotic free-for-all, with AI agents querying sources directly, is a regression that would erase a decade of progress in data warehousing and governance. To solve this machine-scale problem, we must deploy a machine-scale solution. This session casts a vision for the future, exploring why current models are ill-equipped for the AI era. We will introduce the concept of the virtual data engineer—an AI-powered partner designed to augment and accelerate human capabilities on a collaborative platform. Discover how to evolve your team and architecture to turn this challenge into a strategic advantage, ensuring you lead the way through this transformation.

Unstructured text — like customer surveys, policy documents and open-ended feedback — can be a goldmine of insights, but extracting and operationalizing that intelligence has traditionally been a challenge. We'll explore how LLMs are revolutionizing the data engineering pipeline by translating “messy” text into concise, actionable answers. By transforming yes/no questions into actionable data points, LLMs empower businesses to surface trends, drive BI reporting, and fuel predictive modeling with meaningful categorical information. See how our customers eliminated thousands of hours of tedious manual data processing in fields such as insurance and finance.