Smartsheet launches AI server for work management data

Smartsheet has launched a Model Context Protocol server that connects artificial intelligence tools to data in its work management platform. More than 4,000 users adopted the product in its first few days.

The server is designed to let AI assistants access live project and operational data in Smartsheet, including tasks, timelines, risks and dependencies. It is available now with a native integration for Anthropic's Claude, while integrations for ChatGPT and Gemini are expected to follow.

The launch reflects a broader push by software companies to make AI systems do more than generate summaries or answer prompts. In work management, that means linking models to the data and workflows project teams use so they can update records, identify issues and support day-to-day execution.

Customers can use the server with more than 35 AI tools. Early usage has generated 1.74 million total actions since launch, with 48% of actions taken through Claude moving work forward by creating tasks or making updates rather than simply retrieving information.

Early Uptake

Smartsheet is positioning the product as a way to address a common problem in corporate AI deployments: access to structured operational data. Many AI tools can analyse documents or answer general questions, but they often lack direct links to the systems where project plans, dependencies and workflow changes are managed.

That gap has become more visible as businesses try to measure returns on AI spending. Smartsheet cited Gartner data showing that only one in five AI projects delivers return on investment, underscoring how difficult it is for organisations to turn AI budgets into operational results.

Smartsheet says its platform is used by more than 100,000 organisations, creating a large installed base for the new offering. The system is designed to work across an organisation's existing AI tools rather than forcing teams into a single assistant.

"For years, Smartsheet has been where enterprise work happens. With our MCP, customers are now connecting their AI tools directly to the data and systems that run their business," said Pratima Arora, Chief Product and Technology Officer at Smartsheet. "We are moving past the 'chat' phase of AI and into a new era where AI can impact the productivity of teams working on strategic projects."

Cost Focus

Alongside workflow access, Smartsheet is making a cost argument for AI usage. Large language model providers typically charge based on token consumption, and companies have become increasingly alert to how those costs can rise as AI tools are used more widely across an organisation.

Smartsheet says its MCP server compresses and filters data before sending it to AI models, reducing the amount transferred and lowering token-related fees. Lighter data payloads can also improve response times when users query large project portfolios.

Governance is also central to the launch. Every interaction initiated by a human user or by AI is logged and made available for audit, allowing organisations to monitor how AI tools interact with work data, which tools are used and how tokens are consumed.

That audit trail is likely to appeal to companies under pressure to show that AI use is controlled, particularly in sectors where project delivery, compliance and internal approvals depend on clear records of changes and decision-making.

Liz Miller, Vice President and Principal Analyst at Constellation Research, said the commercial test for enterprise AI lies beyond content generation.

"The true ROI of enterprise AI isn't measured by how many questions can be asked or how much text can be generated; that is the output of AI. It lies in how precise decisions, actions and outcomes can become," Miller said. "By integrating AI models, businesses can use AI as a productivity multiplier. Combining AI models with deep organisational data creates a business intelligent advisor, closing a critical visibility gap that can cost large organisations thousands of hours in status meetings and manual risk assessment."

Partner View

The product is also being adopted by partners that use Smartsheet in customer projects. Those partners often work closely on large-scale transformation programmes and can influence how AI tools are integrated into project and operations software.

Julian Weber, Partner at AMX, described the Claude integration as an extension of Smartsheet's existing role in giving customers visibility across complex project data.

"For years, we've relied on Smartsheet to help our customers create transparency across complex, global project data. The native Claude integration builds on that foundation, bringing an AI layer that surfaces trends, risks and critical changes through simple, natural conversation based on live business data," Weber said. "That's a meaningful shift in how teams operate and make decisions. The opportunity ahead is enormous: as we roll this out for our customers, we're transforming how they act on their data at scale, enabling users to create intelligent, real-time insights-easier and faster than ever before."

The MCP server is part of a broader effort by Smartsheet to move AI from individual assistance to organisation-level use inside project and operational systems. Early figures suggest that shift is already under way, with 48% of AI actions resulting in task creation or updates rather than information requests alone.