Bay Area Snowflake User Group

September 28, 2026 · Bay Area

AI Agents for Data Teams

A practical mini-hackathon.

Bring a real data-team problem. Work with others to learn how to use Snowflake’s agentic AI products to find, access and govern data—and build something that adds value to your team.

The September 28 outcome

Leave with something useful.

01

Experience building together

Collaborate with others on Snowflake’s agentic AI products to solve a data-team problem. Learn from how your teammates approach the work.

02

Progress on a real task

Build and run something useful. No problem is too small if solving it adds value to a team or the company.

03

A result you can explain

Show what you got running, explain the value it creates and share what you learned along the way.

Start with your work: find the right data, access it appropriately and govern how it is used. Explore documentation, data quality, lineage, discovery, pipelines or another task you care about with AI agents.

For discussion with Snowflake

Snowflake products for the session.

A starting point for choosing what teams will use on September 28. The product selection, walkthrough and account setup are to be agreed with Snowflake.

Build

Snowflake CoCo

An AI coding agent to help build data workflows and agents.

Formerly Cortex Code.

Connect data & tools

Cortex Agents

Build agents that work with governed data and choose tools to answer questions or carry out tasks.

Use & explore

Snowflake CoWork

A conversational application for working with enterprise data agents.

Snowflake CoWork routes a question to an agent that plans, tracks details, uses tools and reflects. Cortex Search retrieves documents, Cortex Analyst works with structured data, and custom tools perform tasks.
Snowflake’s illustration of how CoWork uses an agent, data and tools. Source: Snowflake’s Cortex Agents guide ↗ · View larger ↗

September 28 · 5–8 PM

One evening. Time to make progress.

5:00 PM
Arrive & find your team

Meet collaborators and choose a data-team task to explore.

5:30 PM
Welcome & a practical starting point

Introduce the challenge, the Snowflake platform and how memory can support AI agents.

5:45 PM
Use Snowflake’s agentic AI products

Ninety minutes to collaborate on a data-team problem, build with the products and get something running.

7:15 PM
Show what works & share what you learned

Record a three-minute demo: the value created, your use of the tools and what actually runs. Share highlights and feedback. Wrap up at 8 PM.

Bring: a laptop and a data-team task you’d like to explore with AI agents. Snowflake account setup will be shared before the session. Work with public, synthetic or approved sample data you can share with your team.

Three criteria

How we’ll review the work.

No problem is too small. Show why it matters and how far you got.

Value Creation

Does it add value to a team?

Show who benefits and what gets better for them or the company.

Innovation

Are you using the tools in new ways?

Explain what you tried and why that approach is useful.

Working

How much did you actually build and get running?

Demonstrate what works and be clear about what remains.

For Snowflake review

Support for September 28.

Help make this a useful working session for the Bay Area community.

Lead sponsor · Agentic AI platform

Snowflake

Learn how to use Snowflake’s agentic AI products with governed data. The session will give teams a practical starting point for finding, accessing and governing data with agents.

Coordinate the product walkthrough, account setup, participant credits and technical support for the evening.

Memory contribution

Cognee

Memory for AI agents, with Dave Nielsen supporting teams as they retain and reuse useful context across tasks.

Explore what to remember, who can reuse it and how it stays current as source information changes.

The working example and session support are to be agreed.

Make the evening work

Event support

Coordinate venue logistics, refreshments and promotion for an evening planned around 75 attendees.

Snowflake leads the session. Account setup, credits, the memory example and event logistics remain to be agreed.

NEXTOCT 9

Los Angeles

Build on September 28

Take the results to DAMA Days.

We expect September 28 to produce practical examples, participant feedback and a format worth building on. A successful session gives us a stronger basis for the proposed October 9 hackathon at Loyola Marymount University.

The next request: Snowflake funding or sponsorship for October 9, alongside AI agent platform access, mentors and a connection to the conference program. This is the follow-on opportunity.

About DAMA Days, October 9 ↗