Small models don’t need more parameters, they need better data. I’ll share how my team built the xLAM family of small action models that punch far above their weight, enabling fast and accurate AI agents deployable anywhere. We’ll explore why high-quality, task-specific data is the ultimate performance driver and how it turns small models into powerful, real-world solutions. You’ll leave with a practical playbook for creating small models that are fast, efficient, and ready to deploy from the edge to the enterprise.

For years, data engineering was a story of predictable "pipelines": move data from point A to point B. But AI just hit the reset button on our entire field. Now, we're all staring into the void, wondering what's next. While the fundamentals haven't changed, data remains challenging in the traditional areas of data governance, data management, and data modeling, which still present challenges. Everything else is up for grabs. This talk will cut through the noise and explore the future of data engineering in an AI-driven world. We'll examine how team structures will evolve, why agentic workflows and real-time systems are becoming non-negotiable, and how our focus must shift from building dashboards and analytics to architecting for automated action. The reset button has been pushed. It's time for us to invent the future of our industry.

Start with a dataset in Motherduck and build a production-ready analytics app using Omni’s semantic model and APIs. We’ll cover practical data modeling techniques, share lessons learned from building AI features, and walk through how to give AI the context it needs to answer questions accurately. You’ll leave with a working app and the skills to build your next one.

Get ready to ingest data and transform it into ready-to-use datasets using Python. We'll share a no-nonsense approach for developing and testing data connectors and transformations locally. Moving to production will be a matter of tweaking your configuration. In the end, you get a simple dataset interface to build dashboards & applications, train predictive models, or create agentic workflows. This workshop includes two guest speakers. Brian teach how to leverage AI IDEs, MCP servers and LLM scaffoldings to create ingestion pipelines. Elvis will show how to interactively define transformations and data quality checks.

Learn to build an autonomous data science agent from scratch using open-source models and modern AI tools. This hands-on workshop will guide you through implementing a ReAct-based agent that can perform end-to-end data analysis tasks, from data cleaning to model training, using natural language reasoning and Python code generation. We'll explore the CodeAct framework, where the agent "thinks" through problems and then generates executable Python code as actions. You'll discover how to safely execute AI-generated code using Together Code Interpreter, creating a modular and maintainable system that can handle complex analytical workflows. Perfect for data scientists, ML engineers, and developers interested in agentic AI, this workshop combines practical implementation with best practices for building reasoning-driven AI assistants. By the end, you'll have a working data science agent and understand the fundamentals of agent architecture design. What you'll learn: ReAct framework implementation Safe code execution in AI systems Agent evaluation and optimization techniques Building transparent, "hackable" AI agents No advanced AI background required, just familiarity with Python and data science concepts.

Everyone's trying to make LLMs "accurate." But the real challenge isn't accuracy — it's context. We'll explore why traditional approaches like evals suites or synthetic question sets fall short, and how successful AI systems are built instead through compounding context over time. Hex enables a new workflow for conversational analytics that grows smarter with every interaction. With Hex's Notebook Agent and Threads, business users define the questions that matter while data teams refine, audit, and operationalize them into durable, trusted workflows. In this model, "tests" aren't written in isolation by data teams — they're defined by the business and operationalized through data workflows. The result is a living system of context — not a static set of prompts or tests — that evolves alongside your organization. Join us for a candid discussion on what's working in production AI systems, and get hands-on building context-aware analytical workflows in Hex!