Snowflake introduced a research preview of Project SnowWork, a new autonomous AI platform built within its data cloud environment, designed to bring outcome-driven, agentic AI capabilities directly to business users across the enterprise. Project SnowWork operates as an AI-driven orchestration capability within Snowflake’s data cloud environment, where agents can access governed enterprise data, interact with systems, and execute workflows.
Currently in research preview, Project SnowWork is designed to enable users to complete complex business tasks using natural language prompts, positioning AI as an active participant in enterprise workflows, the company said.
Sridhar Ramaswamy, CEO of Snowflake, said, “Project SnowWork looks to put secure, data-grounded AI agents on every surface, so business leaders and operators can move from question to action instantly. By elevating AI from experimentation to enterprise-grade autonomous execution, Project SnowWork serves as the secure foundation for how modern enterprises will get work done in the AI era.”
From Data Platforms to Autonomous Work Execution
Snowflake’s latest move reflects a broader shift in enterprise software – from data platforms that provide insights to systems that can act on those insights.
“We are entering the era of the agentic enterprise, ushering in a fundamentally new way to work. This shift is about much more than technology—it’s about unlocking new levels of productivity and efficiency by embedding intelligence directly into the operating fabric of the enterprise,” Ramaswamy said.
Project SnowWork works by deploying secure, data-grounded AI agents that interpret natural language prompts and translate them into executable workflows. Once a user issues a request, the system can:
- Query and retrieve relevant enterprise data
- Apply analysis and business logic
- Synthesize insights into structured outputs
- Trigger downstream actions across connected systems
These agents operate within Snowflake’s governed data environment, ensuring that all actions follow existing access controls and data policies.
This represents what Snowflake describes as “outcome-driven AI,” where the focus moves beyond generating insights to completing work – automating workflows that previously required manual coordination across systems and teams.
Embedding AI Into Everyday Business Processes
Project SnowWork is designed to make AI accessible to business users without requiring deep technical expertise. Through conversational interfaces, users can instruct AI agents to perform tasks such as data analysis, reporting, and operational workflows.
By embedding these capabilities directly into the data cloud environment, Snowflake aims to reduce the gap between data access and action, often referred to as the “last mile” of enterprise AI adoption.
Project SnowWork introduces several core capabilities that define how it operates within enterprise environments:
- Pre-built, persona-specific skills: Role-aware AI profiles for functions such as finance, sales, marketing, and operations that understand workflows, KPIs, and business context.
- Multi-step task completion: The ability to autonomously plan and execute complex workflows, querying data, generating insights, producing deliverables, and recommending next steps within a single interaction.
- Built-in security and governance: Enforcement of Snowflake’s role-based access controls (RBAC), masking policies, audit logging, and data governance rules to ensure AI operates within enterprise security boundaries.
While Snowflake has indicated that Project SnowWork is currently available in a limited research preview, signaling that the platform is still evolving, its introduction signals Snowflake’s push to position its data cloud as an execution layer for enterprise workflows, not just an analytics platform.
What This Means for ERP Insiders
Workflow compression reshapes process ownership.
By combining data access, analysis, and execution in one interaction, SnowWork collapses handoffs across teams and tools, changing how finance close, reporting, and operations workflows are designed and owned. For ERP teams, this means re-evaluating process design, ownership boundaries, and where automation can replace multi-system coordination.
Data platforms move closer to ERP and application workflows.
Snowflake’s approach signals convergence between data, analytics, and application layers, where enterprise data platforms begin orchestrating actions across core systems rather than just reporting on them. ERP leaders should expect tighter integration between data platforms and core applications, with increasing pressure to align data models, APIs, and process logic.
Governance and control models must evolve for autonomous AI.
As AI agents execute tasks using enterprise data, organizations will need to extend existing security, audit, and compliance frameworks to ensure visibility, accountability, and control over AI-driven actions. This will require ERP and IT teams to define new guardrails, approval mechanisms, and audit trails for AI-initiated transactions and decisions.
