At the VB Transform 2026 conference in early August 2026, Asana's Chief Product Officer, Arnab Bose, unveiled a new operating system called Agentic Work Management (AWM). This solution positions AI agents as coachable teammates working alongside humans, rather than just basic one-on-one personal assistants. This is considered Asana's strategic move to thoroughly address the statelessness and lack of information linkage that currently bottleneck enterprise chatbots.
Background & Causes
Enterprise AI development teams frequently face the issue where chatbots answer single prompts well but fail to remember what previous users asked. According to Bose, current chat tools, though easily connected via the Model Context Protocol, operate in a vacuum and fail to create reusable workflows. To overcome this, Asana leveraged its Work Graph architecture, built over 18 years, as the foundation for AWM. This architecture organizes data in a "Pyramid of Clarity," linking individual tasks to company-wide goals. Consequently, AWM's AI agents can access a "shared company brain" to track progress and collaborate seamlessly with human colleagues.
Technical & Technology Analysis
On the technical side, Asana's AWM system addresses three major challenges: data governance, dynamic model routing, and billing structures. To prevent sensitive information leaks from confidential projects (such as M&As), Asana's access control system distinguishes between triggering new memory creation and simply executing tasks. Additionally, AWM automatically routes complex prompts to heavy frontier models like Anthropic's Claude Opus or OpenAI's models, while down-leveling lighter tasks to cheaper ones. To relieve customers of token-counting burdens, Asana charges a flat rate per task completion instead of variable computing costs.
Expert Opinions & Insights
Asana reported that this solution is already live for several major customers, including FedEx and cloud provider CoreWeave. According to Bose, CoreWeave uses AWM and multiple specialized AI agents to automate its new product launch workflows, which previously required complex manual approvals. However, analysts note that Asana faces fierce competition as the very frontier-model providers powering AWM, such as Anthropic and OpenAI, are shipping their own agent products. Facing these "frenemies," Bose confidently argued that Asana's advantage lies in its nearly two decades of user experience and workflow data.
Impact & Future
The Agentic Work Management model promises to reshape how enterprises operate in the era of autonomous AI. Shifting from personal chatbots to AI agents with shared memory and complex workflow automation will significantly reduce repetitive administrative tasks. For the technology and business community, this serves as a valuable, practical blueprint of combining traditional graph databases with large language models to deliver real business value rather than just chasing industry hype.