New AI Tools Automate Complex Workflows
A new wave of AI automation is hitting the enterprise, with Perplexity launching "Computer," an AI that can execute entire multi-step workflows. Meanwhile, project44 debuted an AI agent to autonomously resolve container shipping disruptions.
Perplexity's "Computer" operates as a multi-model orchestration platform, not a single AI. It decomposes a user's goal into a structured graph of tasks, then routes these subtasks to one of 19 different specialized AI models, such as Anthropic's Claude for reasoning and Google's Gemini for deep research. This system allows "Computer" to function as an autonomous digital worker, capable of handling complex projects for hours or even months. The system is designed to be model-agnostic, meaning Perplexity can swap in newer, more capable models as they emerge. Each task assigned by the orchestrator runs in an isolated cloud environment with its own filesystem and browser, a security measure to prevent the issues seen with local-access AI agents. This cloud-first approach contrasts with open-source agents that run directly on a user's machine. The project44 AI Ocean Exceptions Agent is built upon the company's vast logistics data graph, which tracks 1.5 billion shipments annually across 259,000 carriers. This allows the agent to identify potential "rolled container" risks up to 35 hours before official carrier notifications. The tool operates within project44's broader "Multi-Agent Orchestration" framework, which already handles millions of automated carrier communications per year. Upon detecting a likely disruption, the project44 agent autonomously confirms the issue with the carrier, sources alternative voyage options, and presents a structured recommendation to a human analyst for a final decision. This process reduces the time required for roll detection and rebooking readiness from hours to under five minutes. Early deployments of similar AI procurement agents from project44 have shown a 4.1% reduction in freight spend and a 70% decrease in manual coordination efforts. These tools represent a broader enterprise trend toward "agentic AI," where autonomous systems make decisions and execute tasks without constant human intervention. The goal is to move beyond simple automation of repetitive tasks to optimizing complex processes that require reasoning and problem-solving. According to Gartner, 80% of enterprises are expected to be using AI APIs and workflow automation platforms by 2026.
