For global enterprises like the Mark Anthony Group (MAG), managing vast, disparate data sets is a significant operational hurdle. MAG, owner of brands like White Claw, operates across diverse markets, each with unique systems and data sources. Traditional data warehousing struggled to keep pace.
MAG is now shifting from passive business intelligence to what Senior Director of Data, Analytics and AI Sam Wong calls a "generative BI evolution" powered by Snowflake Intelligence. This initiative aims to put conversational data access directly into the hands of business users, transforming MAG into an agentic enterprise.
The company views Snowflake not just as a vendor but as an innovation partner, prioritizing its Secure Data Sharing capabilities in RFPs. This approach has pushed third-party data providers and global vendors away from archaic flat-file transfers towards reliable, real-time data integration.
This standardization has demonstrably reduced total cost of ownership, boosted reliability, and improved data quality by addressing issues at the source. Even MAG's sales and distribution partners benefit, finding access to data significantly easier through Secure Data Sharing compared to traditional SFTP and ETL processes.
Democratizing Data with Custom AI
MAG has deployed a custom-built global enterprise application that acts as a wrapper for the Snowflake Intelligence engine. This pilot application is designed to democratize data across all teams and business units within the group.
Key features of MAG's implementation include:
- Explainability and observability: A crowdsourced system provides context for data sets, explaining the 'why' behind the numbers.
- Mobile and voice integration: Web-enabled and mobile-friendly, allowing executives to query metrics on the go.
- Collaboration through Teams: Integration into Microsoft Teams provides multiple access points, avoiding data silos.
The tool uses Snowflake Intelligence's text-to-SQL capabilities, allowing users to ask complex questions in plain English or via voice commands, bypassing the need to understand underlying SQL code.
Solving the Semantic Challenge
A core challenge in AI deployment is ensuring models understand business-specific language. MAG is addressing this by integrating Snowflake with partners like Ataccama to build a tool-agnostic semantic layer.
By merging its business glossary and data catalog into Snowflake, MAG is tuning its data so the AI accurately interprets internal acronyms and terms across its diverse markets.
This initiative is about driving tangible business outcomes: increased revenue, reduced costs, and improved customer experience.
MAG has already built AI/ML engines for sales performance recommendations and data enrichment, with Snowflake Intelligence expected to expedite decision-making and subsequent actions across the company.
Wong anticipates this will trigger new levels of data utilization, fundamentally changing business processes and workflows toward an agentic enterprise vision.
Expanding to Unstructured Data
MAG is pushing the boundaries by applying Snowflake Intelligence to unstructured data, prototyping systems that allow users to query hundreds of Standard Operating Procedure (SOP) documents and receive referenced instructions.
Leveraging technologies like Document AI, MAG aims to transform every piece of information, whether in a database or a PDF, into an actionable asset.
Wong advises other retail and consumer goods organizations embarking on their AI journey to use specific use cases as their guiding principle.
Focusing on the problem to be solved ensures AI initiatives remain business enablers rather than mere technology exercises.
