Smartsheet has launched vsActual, a standalone AI application that tracks whether company plans are slipping and recommends corrective action. It can draw on connected business apps and, with permission, make changes itself, though human approval is the default.
Watch Desk analysis
What happened
The application is aimed at senior executives and is now in early access, with general availability planned for early 2027. Computerworld reports that vsActual connects to services including Oracle, Salesforce, SAP, Workday and Slack, and uses Smartsheet’s Execution Graph to map relationships between tasks, projects and strategic goals.
Its agents have distinct jobs: an intake agent builds a plan from documents such as presentations and spreadsheets; a planning agent supports capacity planning and scenario modelling; and a drift agent looks for budget overruns, delays and resource conflicts. Other agents monitor data quality and report changes. They can recommend fixes, such as assigning more staff to a delayed project. With relevant permissions, agents can update third-party apps, but human approval is required by default. Pricing has not been disclosed; Smartsheet says billing will be usage-based.
Why it matters
This is a move from organising work to helping executives monitor strategy and act on problems across several systems. The useful distinction is that an agent may do more than flag a red status: it can propose a response and, if granted permission, carry it out.
That makes data quality and oversight central, not decorative. Constellation Research analyst Liz Miller told Computerworld that the accuracy of the system’s outputs will depend on customers’ operational and financial data. The launch also asks Smartsheet to sell to a new audience, beyond its familiar project and operations teams.
Our read
A strategy dashboard that can spot drift and suggest a fix is a more interesting proposition than another dashboard that merely looks concerned. But the permission model deserves a close look: human approval by default is a sensible starting point, while autonomous changes across business systems call for clear limits and a reliable record of what happened.
For now, vsActual is in early access, so the practical questions are how well it handles messy business data and whether its recommendations earn executive trust. “AI-native” is a label; preventing budget creep is the job it will have to do.
What to watch
- Whether early-access customers can show that the agents identify problems accurately.
- What controls and audit trails govern changes made to connected applications.
- Which integrations and pricing terms are available when general release arrives in early 2027.
Discussion spark: Should AI agents be allowed to change live company plans and systems once managers grant permission, or should a person approve every action?
Sources and evidence
- Source update (5 November 2026, 08:48 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.