Agent use cases hiding in document-heavy workflows

Summary

Aaron Levie argues that always-on and workflow-triggered agents will displace chat-based patterns as the primary mode of agent use; read the original tweet. His examples are document-heavy, system-connected workflows: contract review feeding into Linear, client onboarding, invoice processing, M&A due diligence, and data extraction. The infrastructure needed to support these at scale includes long-running execution, safe code execution, tool access, compute sandboxes, and cross-system connectivity.

Key Claims

• Background and workflow-triggered agents will become the dominant source of token use, not chat-based interactions.
• Candidate workflows include contract review, client onboarding, invoice processing, M&A due diligence, and data extraction pipelines.
• The Claude Managed Agents example connects Box and Linear: contract intake triggers review, which populates a Linear task with critical information.
• Deploying background agents at scale requires long-running execution, accurate results, safe code execution, tool access, compute sandboxes, and system connectivity.

Quotes

• "In Claude's new Managed Agents feature, in a couple minutes you can wire up an agent that can read contracts when they come into Box to review them, and then assign a task in Linear with the critical information from the contract."
• "But this could have been any workflow, like reviewing documents for client onboarding, invoice processing, M&A due-diligence, data extraction pipelines, and millions of other use-cases."
• "Agents have the ability to execute code safely, leverage tools, access a compute sandbox, and connect across systems is clearly the architecture of the future."