Streaming to Lakehouse: How Launchmetrics Simplified Pipelines, Built Data Trust and Adopted Genie
Overview
| Experience | In Person |
|---|---|
| Track | Data Engineering & Streaming |
| Industry | Enterprise Technology, Communications - Media & Entertainment |
| Technologies | AI/BI, Delta Sharing, Unity Catalog |
| Skill Level | Beginner |
At Launchmetrics, a leader in Brand Performance for fashion, luxury and beauty, our AWS real-time streaming architecture delivered low latency but created growing complexity, data inconsistencies and rising infrastructure costs.We shifted to a batch-first Databricks Lakehouse built on Delta Lake, using medallion architecture and Unity Catalog for governance. The impact exceeded expectations: fewer pipelines, predictable costs, safe historical recomputation, and stronger data reliability—restoring trust across product, commercial and leadership teams.Most importantly, this foundation enabled company-wide adoption of AI/BI Genie. Today, non-technical stakeholders query the lakehouse in natural language, reducing dependency on the data team and accelerating decisions.In this session, we’ll share how we evaluated real-time vs. batch trade-offs, executed the migration safely and designed a “Genie-ready” lakehouse that delivers measurable business value.
Session Speakers
Pau Montero Pares
/CTO
Launchmetrics
Albert Vila
/VP Data Software Development
Launchmetrics