AI Tools & Products
Jun 25, 2026
MIT and Microsoft Develop System to Enhance Efficiency of AI Workflows
Jun 25, 2026
AI Summary
Researchers from MIT and Microsoft have created a system called Murakkab that optimizes the design and implementation of AI-powered workflows. This system reduces computational needs and energy consumption while maintaining performance, addressing concerns about resource allocation in cloud computing.
- Agentic workflows are AI systems that combine multiple models and tools to perform complex tasks, such as video analysis.
- Traditional workflows often lead to inefficiencies, resulting in wasted computation, energy, and costs.
- The new system, Murakkab, allows developers to describe workflows in high-level terms, automatically optimizing the selection of models and tools, as well as hardware configurations.
- Murakkab adapts configurations based on user priorities, such as cost and speed, and can dynamically adjust to new models or hardware.
- In tests, Murakkab reduced computational requirements by about 65% and energy consumption by approximately 73% compared to traditional methods, while also lowering costs significantly.
- The system provides cloud providers with better visibility into workloads, enabling more efficient resource sharing.
- Future plans include expanding Murakkab to handle more complex workflows and larger computing clusters, aiming for greater resource optimization in cloud platforms.
- The research was supported by the Semiconductor Research Corporation and the U.S. Defense Advanced Research Projects Agency.
ai agentsenergy efficiencyworkflow optimizationsystem designmultistep workflows