Gimlet Labs has launched a platform that disaggregates AI workloads and intelligently maps components to appropriately sized hardware. This approach aims to streamline the computing-intensive processes, addressing the underutilisation of powerful chips like Nvidia's GPUs. The software startup's platform can decompose AI workloads into different components and map them to the most efficient device. Gimlet Labs' platform supports entire agentic applications, spanning multiple models, custom code, and custom data sources.
Gimlet Labs' technology allows developers to deploy on optimal hardware without rewriting applications. The company's platform is deployed at AI-native and Fortune 500 companies, powering workloads that span multi-generation and multi-vendor hardware. Gimlet Labs offers its stack as an on-prem solution for datacenters, as well as a hosted developer platform. The company also raised $12 million in seed funding led by Factory.
The company's software intelligently allocates AI tasks across various chip architectures, including Intel, AMD and Nvidia, to improve efficiency. Gimlet Labs is positioning itself as the VMware of the AI era by enabling cost-effective handling of workloads. Gimlet Labs's serverless inference allows users to run simple to complex multi-agent systems with custom logic and data sources.
