Paul Winsor, Head of Retail EMEA at Snowflake, has spent 40 years in the retail industry — beginning his career with UK supermarket Sainsbury's. According to Winsor, the conversation with retailers has fundamentally shifted: "2026 has really changed in terms of retailers talking much more about the agentic era." Today's Snowflake retail customers are laser-focused on deepening customer understanding to drive meaningful engagement.

Sainsbury's stands as a prime example. With 18 million people enrolled in its Nectar loyalty card scheme, the supermarket now delivers hyper-personalised weekly offers — every Friday, each member receives ten tailored promotions based on their individual shopping history. That programme drives more than 260 million personalised price and points promotions every week, made possible by consolidating all customer data onto a single Snowflake platform.

Sportswear brand ASICS takes a similarly data-rich approach through its Runkeeper app. By aggregating performance data from GPS devices, smartphones, and wearables such as Garmin, Fitbit, and Apple Watch — including location, pace, and distance — ASICS brings all of that data together in Snowflake to deliver personalised content and training insights to its running community.

Personalisation at scale is only the beginning. The more transformative shift happening right now is the move toward live, agentic intelligence — AI that acts in real time, not in hindsight. Snowflake EVP of Product Christian Kleinerman brought Jung Suh, Head of Digital Commerce at Samsung, onstage to demonstrate exactly this.

Samsung is using Snowflake CoWork — the conversational assistant that lets employees query their company data in natural language — to monitor its Galaxy S26 series launch in real time. Suh explained that the traditional analytics workflow was already too slow: by the time a team surfaced why conversion had dropped in a region, the window of action had already closed.

The solution is Samsung's internally built Shoppers Inside Action agent. Rather than simply returning raw numbers, the agent plans a series of analytical steps, reconciles multiple data signals, and delivers a synthesised answer — compressing hours of analyst work into seconds. As Suh put it, those seconds matter when you are troubleshooting a live product launch.

Walking the summit floor, one message was unmistakable: enterprise AI has moved decisively past experimentation. The question is no longer whether to deploy AI, but how to do so safely, quickly, and with measurable business outcomes. The answer, according to Snowflake's closest partners, lies in a deeply integrated ecosystem built around Snowflake Cortex — turning raw data into agentic workflows.

At the foundational infrastructure layer sits the multi-billion-dollar alliance between Snowflake and AWS. Mona Chadha, Director of Strategic Partnerships at AWS, highlighted that this partnership has already generated US$7 billion in joint contract value. Snowflake's AI architecture is built directly on Amazon Bedrock, using AWS Graviton high-performance compute to power its native AI capabilities.

Sigma takes a different approach, targeting the non-technical business user. Rather than forcing teams to learn complex code, Sigma acts as an intuitive interface sitting directly on top of the data warehouse, ensuring that strict data governance and security boundaries are never compromised. As the Sigma team put it: "Snowflake provides the building blocks — Sigma is the easy way to work with those blocks, always keeping within Snowflake."

For teams building complex data pipelines and predictive applications, Dataiku provides critical visual orchestration. The company's recently announced Dataiku Cobuild for Snowflake enables joint customers — such as Air Canada, which uses the combined tech stack for advanced marketing modelling — to build and deploy AI workflows more efficiently. Dataiku emphasised at the event that the most successful enterprises are those working backwards from the business problem rather than chasing the latest technology trend.

Rounding out the ecosystem, systems integrator Tredence focuses on deploying AI into specific industry verticals. Rather than pursuing fully autonomous systems, Tredence advocates for a balanced human-in-the-loop architecture — training highly specialised AI personas on data pulled directly from Snowflake, keeping humans both in and around the decision-making process.

Key Takeaways

• →Enterprise AI has shifted from experimentation to execution — organisations are no longer asking if they should deploy AI agents, but how to do so safely and at scale.
• →Sainsbury's runs over 260 million personalised promotions per week by unifying 18 million loyalty customers' data on Snowflake, setting a benchmark for retail hyper-personalisation.
• →Samsung's Shoppers Inside Action agent, built on Snowflake CoWork, compresses hours of launch analytics into seconds — enabling real-time responses during flagship product rollouts.
• →The Snowflake–AWS alliance has generated over US$7 billion in joint contract value, with Snowflake's AI architecture natively integrated with Amazon Bedrock and AWS Graviton compute.
• →Sigma, Dataiku, and Tredence are each solving distinct layers of the enterprise AI stack — from intuitive no-code analytics interfaces to advanced pipeline orchestration and industry-specific AI personas.
• →The most successful AI deployments are being built backwards from business outcomes — not forwards from technology — with human oversight remaining central to responsible agentic architectures.