Snowflake unveils Project SnowWork for autonomous AI work

Snowflake has introduced Project SnowWork in research preview, describing it as an autonomous AI platform for business users to complete multi-step work from a desktop interface.

It aims to close a common gap in enterprise analytics: insights often sit in dashboards or reports, while teams still rely on manual processes to turn findings into actions. Project SnowWork centres on conversational prompts and workflow execution across governed data in Snowflake.

The preview is limited to a small set of customers. Snowflake has not disclosed pricing, a timeline for broader availability, or the organisations in the initial cohort.

Project SnowWork is designed to plan, analyse, and execute workflows. Examples include assembling forecast materials for board reporting, producing a spreadsheet that flags churn risk, and identifying supply chain bottlenecks. Snowflake says it can complete tasks end to end, rather than returning an answer that users still need to operationalise.

"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," said Sridhar Ramaswamy, Chief Executive Officer, Snowflake.

"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 insight to action

The launch comes as the software industry pushes towards so-called agentic systems, where AI takes a more active role in carrying out work. Snowflake frames the shift as moving beyond question answering to connecting intelligence, applications, and enterprise context in ways that drive decisions and follow-on tasks.

The SnowWork desktop experience is intended to expose Snowflake's data platform and AI tooling to non-technical users. Snowflake says the system plans workflows and executes steps across data, analysis, and deliverable creation, while also proposing recommended actions and prioritised next steps for a given role.

Snowflake is also differentiating the product from general-purpose agents by emphasising governed metrics, shared business definitions, security controls, and auditability. These long-standing enterprise requirements have become more visible as companies consider allowing AI to access sensitive data and trigger changes across business systems.

Security and controls

Project SnowWork automatically enforces Snowflake role-based access controls, along with masking policies, audit logging, and governance rules, according to the company. This keeps AI actions within the same permissioning framework used for data access. Snowflake also highlights cross-cloud interoperability, reflecting the reality that many large organisations store and process data across multiple environments.

Snowflake highlights three functional areas. The first is pre-built, persona-specific skills: role-aware profiles for finance, sales, marketing, and operations, designed around common terminology, workflows, and key performance indicators.

The second is multi-step task completion, where the platform can query data, run analysis, synthesise results, and generate structured deliverables in a single interaction. Snowflake positions this as a way to shorten reporting cycles and reduce manual coordination across tools.

The third is built-in security and access controls, which Snowflake argues will be critical for organisations that need auditable processes and strict data governance.

Industry analyst Sanjeev Mohan, Principal at SanjMo, said the product reflects a shift in enterprise AI spending from analysis towards execution inside day-to-day workflows.

"Enterprises have invested heavily in data platforms and AI, yet the last mile of translating governed data into everyday business outcomes remains largely manual," Mohan said.

"Project SnowWork represents a meaningful shift from AI as an analytical tool to AI as an execution layer embedded directly into enterprise workflows. By grounding autonomous task execution in trusted, governed Snowflake data, shared business definitions, and cross-cloud and cross-domain interoperability, the company is extending its platform from a system of insight to a system of action, which is where measurable business value is ultimately realized," he said.

Product context

SnowWork is part of a broader Snowflake portfolio focused on enterprise AI on governed data. Snowflake Intelligence, which Snowflake describes as an enterprise intelligence agent, focuses on answering business questions and providing verifiable responses within Snowflake's environment. Project SnowWork is positioned as the next step, carrying insights into the execution of multi-step tasks on top of Snowflake data.

Snowflake also pointed to Cortex Code, a coding agent aimed at technical users working on data engineering, analytics, machine learning, and agent-building tasks. Snowflake describes it as a way to generate code and orchestrate workflows, with an emphasis on moving from prototype to deployment.

Snowflake reported more than 13,300 customers globally. It says the research preview of Project SnowWork is the first public view of its direction for autonomous work inside enterprises, with the initial rollout limited to a select group of customers.