Gimlet Labs Inc., an innovative artificial intelligence startup, has secured $12 million in funding, led by Factory and notable investors including Intel’s Lip-Bu Tan and Figma’s Dylan Field. The company’s breakthrough software platform simplifies the process of porting AI models across various graphics chips without requiring any code. This technology not only enhances cost efficiency by adapting models to superior chips but also optimizes performance by disaggregating AI agents into components tailored for specific processors. By mapping tasks to the most appropriate accelerators—such as high-throughput GPUs for compute-bound tasks—Gimlet significantly reduces inference costs. Their custom compiler further fine-tunes AI models using advanced techniques like operator fusion to enhance speed. With existing revenues in the eight-figure range and a clientele that includes Fortune 500 companies, Gimlet Labs is poised to revolutionize AI deployment in both on-premises and cloud environments.
Keywords: AI startup, funding, AI model optimization, processor compatibility, GPU performance, inference costs, Fortune 500 clients.
