How Tipalti Uses Snowflake Cortex AI to Enable 600+ Ad-Hoc Analyses and 5x Query Capacity for $75B Payments Platform

Tipalti processes $75 billion in annual payments across global payables workflows for thousands of high-growth businesses. At that scale, data engineering — both for internal analytics and customer-facing product features — needs to be fast, governed, and capable of surfacing AI-powered insights without requiring data teams to manage complex ETL pipelines or external AI integrations.

Migrating to Snowflake AI Data Cloud on AWS gave Tipalti a foundation for centralizing data governance and query performance. The team deployed Snowflake Snowpark for custom processing functions, Cortex AI with LLM capabilities for natural language understanding across payment data, Cortex Semantics Models for embedding and semantic search, and Cortex Search for fast retrieval-augmented queries against structured and unstructured data.

The core architecture benefit, as described by Michael Naor, Senior Data Engineer at Tipalti, is bringing AI to the data rather than the reverse: "Snowflake has enabled us to bring AI capabilities to our data, rather than moving data to the AI." This eliminates data export overhead, reduces latency, and preserves the governance controls that a payments platform requires for compliance.

The measurable outcomes spanned both performance and product breadth. Simultaneous query capacity increased 5x, removing bottlenecks for analytics teams running concurrent reporting. 600+ ad-hoc analyses were deployed across teams. More than 50 new features were shipped across 10+ user groups, with Cortex AI capabilities enabling product teams to build LLM-powered features directly on payment data. As Yair Taito, Director of Data Engineering, described: "The insights are now always there for our people...they just get answers."