talk-data.com

Agentic AI transforms how data work gets done. By equipping LLMs with governed data and enterprise tools, it can plan, verify, and execute complex workflows — turning months of pipeline work into days while preserving quality and governance.

In this session, former Snowflake and current Genesis engineer, Michael Rainey will share why the founders — former Snowflake executives Matt Glickman and Justin Langseth built Genesis, the first secure, containerized AI agents designed to run natively in the Snowflake AI Data Cloud.

Genesis captures tribal knowledge, proposes mappings, generates dbt-ready pipelines, and prevents downstream breakage — enabling teams to accelerate transformation without adding technical debt.

You’ll learn:

• How Genesis leverages Snowflake’s secure, auditable environment to scale agents safely.
• Architecture choices that preserve institutional knowledge and reduce operational risk.
• Real use cases beyond analysis — from accelerating migrations to delivering production-grade pipelines faster.

The session concludes with a live demo of Genesis inside Snowflake: watch a messy source system converted into a ready-to-run pipeline with lineage, tests, and human-in-the-loop approvals — in minutes, not months.

What We Will Cover

• What is Agentic AI: Core concepts, why agents are not chatbots, and where agents fit in data and workflows.
• What is Snowflake: Services that matter for agents, including governance, compute isolation, Snowpark Container Services, and observability.
• Why Genesis on Snowflake: Data gravity, security boundaries, cost and governance controls, and native execution patterns.
• Getting up and running: Prerequisites, deploying Genesis in Snowpark Container Services, roles and warehouses, connecting sources, enabling lineage and tests, and setting approval steps.
• Use cases and product features: Tribal knowledge capture, schema and mapping proposals, autonomous pipeline generation, impact analysis to prevent breakage, lineage, testing, and approvals.
• Live product demo: End-to-end build from messy source to production pipeline inside Snowflake.

Who Should Attend Data engineers, analytics engineers, platform teams, and architects evaluating governed agents.

Takeaways

• A clear mental model for Agentic AI on Snowflake
• A practical checklist and quick start guide for deploying Genesis natively
• Patterns to reduce technical debt and improve reliability beyond ad hoc analysis

Speakers Michael Rainey

The rise of machine learning has placed a premium on finding new sources of data to fuel predictive models. But acquiring external data is often expensive and many data sets are rife with errors and difficult to combine with internal data. But that’s going to change in 2020.

To help us understand the scale, scope, and dimensions of emerging data marketplaces is Justin Langseth, one of the visionaries in our space. Justin is a VP at Snowflake responsible for the Snowflake Data Exchange.  Prior to Snowflake, Justin was the technical founder and CEO/CTO of 5 data technology startups: Claraview (sold to Teradata), Zoomdata (sold to Logi Analytics), Clarabridge, Strategy.com, and Augaroo. He has 25 years of experience in business intelligence, natural language processing, big data, and AI.