Impala AI emerges from stealth with $11M Seed to slash the cost of running AI models
The startup is targeting growing demand for high-performance, cost-effective AI infrastructure.
Impala AI, an Israeli startup building infrastructure for enterprise-scale artificial intelligence, has emerged from stealth with an $11 million Seed round led by Viola Ventures and NFX. The company says it has developed a new platform for large language model (LLM) inference, the process of running trained models, aimed at cutting costs while giving enterprises more control and scalability.
Full list of Israeli high-tech funding rounds in 2025
The investment will fund team expansion, product development, and customer adoption among large enterprises increasingly strained by the high cost of running AI systems.
Led by CEO Noam Salinger, a former executive at Granulate (acquired by Intel), and CTO Boaz Touitou, Impala is developing what it calls an “AI stack for inference at unlimited scale.” Its proprietary inference engine enables companies to run LLMs directly within their own virtual private cloud (VPC), allowing for a serverless experience while maintaining full control over data and infrastructure.
According to Impala, its approach can reduce costs by up to 13 times per token compared with existing platforms, without imposing rate limits or reliability trade-offs. The company’s first use case will focus on data processing.
According to recent market forecasts, the AI inference sector is expected to reach $106 billion by 2025 and more than $250 billion by 2030, eclipsing training as the central economic challenge for enterprises.
Related articles:
Impala aims to meet that demand by addressing the supply constraints of GPUs and offering an infrastructure layer that scales seamlessly across clouds and regions. Its system deploys directly inside a customer’s environment, allowing organizations to maintain compliance and control while avoiding the data-handling risks that come with external inference providers.
“We’re at the dawn of a new era in AI, where inference will be the driving force behind the next wave of innovation,” Salinger said. “Our mission is to build the infrastructure that powers AI at real-world scale, smarter, faster, and more cost-effectively than ever before.”
