Smartsheet vsActual: AI Agents Track Company Strategy and Suggest Fixes

What happened

Work management software vendor Smartsheet has launched vsActual, a standalone "AI-native" application that helps executives oversee company-wide strategies. Its AI agents monitor progress, flag problems and recommend fixes. Computerworld reports that the product was announced on Wednesday at Smartsheet's Engage event, and that it sits apart from Smartsheet's core work management software.

vsActual pulls data from a range of business systems, including Oracle, Salesforce, SAP and Workday, as well as work management tools such as Smartsheet itself. Its agents can act on natural language prompts from users or run automatically in the background.

The product includes several specialised agents. An intake agent works from existing material, such as presentation decks and spreadsheets, to build a "plan of record" covering goals, timelines, and budget and capacity limits. A planning agent handles capacity planning and scenario modelling, and can be invoked when a plan needs to change. A further drift agent is part of the set, though the available reporting does not give its full remit.

Why it matters

The significance here is less about another dashboard and more about how strategy is maintained. Traditionally, a strategic plan is a document that is reviewed periodically and quickly falls out of step with reality. By turning existing decks and spreadsheets into a structured, machine-readable plan of record, and then keeping that plan under continuous watch, vsActual points towards strategy as a live operating system rather than an annual exercise.

For leaders in digital transformation, the combination of cross-system data access and agents that can rerun scenarios when conditions shift could change how quickly organisations respond to resourcing pressures. The operating-model question becomes who owns the plan of record, how often it is challenged, and which decisions the agents are permitted to recommend versus execute.

The Renascence take

Most coverage will treat this as a question of whether AI can spot when a project drifts off course. The more interesting risk is what the system is anchored to. A plan of record built from existing decks inherits every unexamined assumption those decks contain, and monitoring agents will then measure progress against those assumptions rather than against current customer needs.

The first plan becomes the reference point, and reference points are sticky: this is anchoring at organisational scale. An agent that flags deviations from a flawed baseline will simply make a flawed strategy run more efficiently. Customer-obsessed operators should make assumptions explicit before ingesting any legacy document, and schedule regular human challenges to the plan itself, not just to its progress. Treat the agent as a tool for testing strategy, not for defending it.