Earlier in June, the Snowplow team joined thousands of data leaders, engineers, and AI practitioners at Snowflake Summit 2025 in San Francisco. OpenAI CEO Sam Altman delivered a fascinating keynote, with much of the conversation focused on the explosion of AI capabilities and agentic systems. But the bigger takeaway for us was this: 

The Snowflake AI Data Cloud is evolving from analytical insights into real-time customer intelligence—and it's increasingly powering customer-facing AI experiences.

Over the course of the conference, we saw the launch of Cortex AI SQL - a powerful new tool that allows you to ask questions about documents, images, and text data using plain English. 

We saw the release of Data Science Agent - an intelligent companion that automates machine learning workflows, letting companies deploy AI models weeks faster without needing specialized data science expertise. 

What Snowflake made clear is that the modern AI Data Cloud is being rearchitected. It’s being primed for AI-native use cases, real-time personalization, and customer-intelligent applications. 

Snowplow was proud to be part of several major product showcases and partner conversations throughout the week, including the introduction of Snowplow Signals—our real-time customer intelligence system that gives AI apps instant access to deep customer context, enabling hyper-personalized experiences without cold starts. 

Here's a closer look at what we announced, what we learned, and how Snowplow is helping teams unlock customer-intelligent AI in the Snowflake ecosystem.

Featured: Snowplow Signals, A Real-Time Customer Intelligence System

In his session on "Powering Tomorrow's Customer Experiences with Cortex AI," Yali Sassoon, Snowplow CTO & Co-founder, highlighted Snowplow Signals. 

Signals provides applications with access to deep, real-time, trustworthy customer context, making it easier to hyper-personalize user journeys and equip AI agents to overcome the "cold start problem."

Yali outlined the three key takeaways needed to power AI-Agents:

1. AI Agents Need Customer Perception to Be Compelling
Most customer-facing agents today operate in isolation without access to customer data, creating "split brain" experiences. Snowplow Signals gives agents both short-term working memory (real-time behavior) and long-term memory (historical context) to deliver truly personalized interactions.

2. Enterprises Can Build Superior Agent Experiences
While AI-native startups threaten to disintermediate customer relationships, established brands have unique advantages: domain expertise, customer expertise, and the data foundation to build customer-intelligent agents that deeply understand their users.

3. Real-Time Customer Intelligence Transforms Support Models
Instead of customers filing support tickets elsewhere, agents embedded in products can observe user struggles and proactively offer help in real-time—turning reactive support into proactive assistance within the application itself.

Snowplow Signals + Snowflake's AI Capabilities: Powering Customer-Intelligent Agents

At Snowflake Summit, Snowplow’s Co-founder and CTO, Yali Sassoon, demonstrated how the combination of Snowflake's AI infrastructure and Snowplow Signals enables enterprises to build customer-facing agents that deeply understand their users—addressing the critical gap in today's AI applications.

As Yali explained in his session on "Powering Tomorrow's Customer Experiences," most customer-facing agents today lack customer perception. They operate in isolation without access to customer data, creating "split brain" experiences where agents can't see what customers are doing on the main site.

Snowplow Signals solves this by providing agents with two types of customer memory:

Short-term Working Memory: Real-time visibility into current customer behavior and the ability to infer intent from actions.

Long-term Memory: Access to full customer history and preferences, using Snowflake Cortex AI to surface relevant data subsets and contextual information.

This customer perception enables agents to:

• Recognize returning customers and understand their preferences
• Proactively intervene when customers struggle (like offering sizing help)
• Provide context-aware recommendations based on session data and historical behavior
• Transform support models by solving problems in real-time within applications

Snowflake's AI Infrastructure Announcements

Snowflake's Summit announcements create powerful synergies with Snowplow Signals:

Cortex AI SQL + Signals: Companies can build sophisticated customer intelligence pipelines using familiar SQL queries on documents and behavioral data, then instantly activate these insights through Signals to power real-time personalized experiences. As a result, you can turn complex data analysis into immediate customer action without requiring separate AI infrastructure.

Data Science Agent + Signals: Data teams can automatically generate ML models for customer segmentation and personalization using natural language prompts. You can then deploy these models instantly through Signals to production applications—reducing time from experimental insight to live customer experience from months to days.

Snowflake Intelligence + Signals: Business analysts can ask Snowflake Intelligence: "Why are customers churning in our mobile app?" in plain English and get answers enriched with live user session data. You can then automatically trigger retention campaigns or personalized interventions—bridging the gap between business questions and immediate customer action.

Horizon Catalog + Signals: Organizations can confidently deploy customer-facing AI agents knowing their behavioral data pipeline meets enterprise security and compliance standards. This enables the rapid scaling of personalized experiences while maintaining data governance—removing the biggest barrier to AI adoption in regulated industries.

Looking Ahead: Building Customer-Intelligent Infrastructure

One message came through loud and clear at Snowflake Summit 2025: companies are moving from analyzing customer data to acting on it in real-time. The modern data stack is evolving beyond analytics to become the foundation for AI-native customer experiences.

As AI agents become central to customer interactions, businesses need customer intelligence that drives immediate action. At Snowplow, we're seeing this transformation across four key areas:

First, AI systems need structured customer context to overcome the cold start problem. We deliver behavioral data that gives AI agents the full customer story—intent, session context, and journey stage—not just raw events.

Second, real-time customer intelligence enables immediate personalization. Companies using Snowplow with Snowflake can trigger interventions, dynamic experiences, and proactive assistance the moment customers signal intent or need.

Third, governed customer data builds trust in AI deployment. With transparent schemas and validated pipelines, teams can confidently launch customer-facing AI knowing their data foundation meets enterprise standards.

Finally, we're empowering teams to build custom customer intelligence rather than relying on generic AI tools. Leading companies are taking ownership of their customer intelligence strategy to create differentiated experiences.

‍

Snowplow Signals is available to select partners now, with general availability in Q3 2025.

Explore Snowplow Signals or get in touch to learn how we're helping Snowflake customers unlock real-time, AI-ready customer data infrastructure.