Serverless AI infrastructure platform company Cerebrium has secured $8.5-million in seed funding, which will allow the team to invest in new features of its platform and meet growing enterprise demand.

The company's platform is purposefully built for high-performance, real-time multimodal AI applications, such as voice agents, large language model fine-tuning, video models and large-scale data analytics use cases.

Founded in Cape Town and headquartered in New York City, in the US, the company builds and scales multimodal AI applications without the traditional complexity or cost.

The seed funding round was led by Internet services company Google's AI venture fund Gradient with participation from seed funding companies Y Combinator and Authentic Ventures, and several strategic angel investors and operators.

The local founders Michael Louis, who is also the company’s CEO, and Jonathan Irwin are pleased to have secured backing from one of the most influential AI investors in the world, Cerebrium says.

Louis and Irwin founded Cerebrium after struggling to build their own AI-driven products.

“Tooling was fragmented, there was an education gap between theory and production, the unit economics didn’t make sense and development cycles took months.

“We built Cerebrium so engineers can focus on building AI products that users love with real business impact, instead of hiring an infrastructure team, racking up six-figure cloud bills or worrying about security and compliance,” says Louis.

“AI is changing the world and we want Cerebrium to be the platform powering it,” he says.

“It is incredible what the Cerebrium team has pulled off with such a small group. They’re powering some of the most advanced AI voice and video applications at scale and we believe specialised infrastructure that scales elastically will be essential as real-time AI becomes core to customer experiences,” says Gradient partner Eylul Kayin.

“[At AI research company Tavus], we run a range of real-time audio and video models, and performance is everything. We tried several solutions, but Cerebrium consistently delivered the speed and reliability we needed without the overhead.

“Even as we’ve scaled rapidly and gone viral, they’ve kept up with our compute demands and delivered the stability we rely on. It has become a core part of our infrastructure,” says Tavus senior machine learning engineer Roey Paz-Priel.

Further, while Cerebrium is known for its serverless graphics processing unit infrastructure, it also offers the ability to do batching, multi-region deployments and large-scale data processing, Cerebrium says.

This enables teams to run compute-intensive workloads with minimal setup, scale elastically and only pay for what they use, without the complexity of managing infrastructure while adhering to strict security and data residency requirements.