The Last Mile of Data: Bridging the Gap Between Warehouse and Business

Hosted by Intercom’s Paul Vickers and Patrick Lindstrom

Join us at the Intercom Dublin office for an evening focused on the "Last Mile" of data engineering—turning solid infrastructure into immediate business value.
Over the last two years, Intercom migrated to Snowflake and rebuilt our data foundations from the ground up, creating a trusted, high-quality layer of core data models. But we realized that great data is only useful if people can actually use it.
In this session, the Intercom Data team will share how we are moving up the value chain—going beyond static dashboards to build intelligent, interactive tools that empower our GTM teams.

The Agenda:

Breaking the SQL Barrier: Scaling Insights with Claude & Snowflake
Even with perfect data models, the "SQL Wall" often forces business leaders to wait in a queue for answers. We’ll demonstrate how we used Claude Code and Snowflake MCPs to remove this bottleneck, allowing non-technical users to query our Data Warehouse directly and create sharable artifacts of their analysis. See how we combined strict governance with an AI interface to turn "data tickets" into instant, self-serve answers.

From Insight to Action: Building 'Cockpit' on Snowflake Container Services
While chat enables exploration, some workflows require action. We will pull back the curtain on Cockpit, a full-stack application built by our Data Science team to help Relationship Managers optimize customer outcomes. Learn how we used AI-assisted coding ("vibe coding") to build the app and Snowpark Container Services to deploy it—proving that Data teams can build powerful software, not just reports.

From Idea to Real-Time Fraud Detection in Snowflake ML
In this session we’ll build a real-time fraud detection service end-to-end inside Snowflake. Starting from a natural language idea, we’ll use AI-assisted development to generate synthetic transaction data, explore patterns, and train an ML fraud model with Snowflake ML. We’ll then deploy it to Snowpark Container Services and expose a low-latency REST endpoint for live transaction scoring. You’ll leave with a practical pattern you can reuse for other real-time ML use cases such as credit risk scoring, real-time recommendations, or anomaly detection on operational data in Snowflake.

Location
Intercom office (near Stephens Green)

Event Schedule

• 6:00 PM: Networking with pizza and drinks
• 6:30 PM: Breaking the SQL Barrier: Scaling Insights with Claude & Snowflake, Intercom
• 7 pm From Insight to Action: Building 'Cockpit' on Snowflake Container Services, Intercom
• 7.30 pm 10 minute break
• 7.40 pm From Idea to Real-Time Fraud Detection in Snowflake ML, Max Zito, Snowflake
• 8.10 pm Chats with drinks
• From 9 pm More chats in a nearby pub