This strategic funding underscores our commitment to supporting visionary founders who are building the essential systems powering the future of enterprise AI. As part of this round, Maxitech brings not only capital but also access to our extensive enterprise network, helping Cerebrium scale its transformative technology into mission-critical production environments across industries.

Meet Cerebrium: Serverless Infrastructure for Real-Time Voice and Video AI

Cerebrium is a serverless AI infrastructure platform purpose-built for the new generation of voice and video AI applications. As demand accelerates for real-time, human-like AI interactions—from voice assistants to automated video generation—developers face enormous complexity in delivering these experiences at scale.

Building production-ready voice or video AI systems is far from trivial. It requires chaining multiple heavyweight models (speech recognition, large language models, text-to-speech, video generation) while maintaining ultra-low latency and managing GPU resources effectively. Traditional cloud services force teams to either invest heavily in bespoke infrastructure engineering or accept high costs and poor performance.

Cerebrium solves this by offering serverless GPU infrastructure tailored to AI workloads. Developers can deploy, scale, and operate AI models without managing servers, clusters, or autoscaling logic themselves. The platform handles orchestration, scaling, and cold start optimizations automatically—enabling seamless, cost-efficient deployment even for complex AI pipelines.

This approach is already powering some of the world’s fastest conversational AI systems, such as voice bots delivering ~500 ms response times, and high-resolution AI video generation features that would be otherwise unachievable for lean teams. Cerebrium’s customers span ambitious startups and enterprise innovators alike, demonstrating the broad demand for truly production-grade AI infrastructure.

Why We Invested

At Maxitech, we specialize in investing in AI companies that transform enterprise operations and strategies. Our mission is to bridge the gap between cutting-edge AI research and real-world business impact by backing the foundational technologies that enable scalable, reliable enterprise deployment. Cerebrium embodies this mission perfectly.

The voice and video AI sector is experiencing explosive growth. In 2024 alone, voice AI startup funding surged 8×, exceeding $2.1 billion, reflecting the demand for more natural, frictionless interfaces between humans and machines. Enterprises are rapidly adopting generative video for training, marketing, and communications. Yet while powerful models exist, delivering them with low latency and cost efficiency at scale remains a major bottleneck for adoption.

Cerebrium tackles this problem at its root by abstracting away infrastructure complexity. With Cerebrium, even small teams can deploy sophisticated AI features that were once the domain of well-funded giants with dedicated ML Ops teams. This democratizes access to advanced AI capabilities, opening the door for a broader wave of innovation in enterprise software.

As Burak Arık, Founder of Maxitech, explains:

“We’re committed to supporting the infrastructure that will power the next generation of AI applications. Cerebrium’s platform stands out for its ability to make high-performance voice and video AI accessible to any developer team. This is exactly the kind of foundational technology we want to see adopted across our enterprise network—technology that solves real deployment challenges and enables seamless, production-ready AI experiences.”

By investing in Cerebrium, Maxitech is not only supporting an outstanding team and product but also advancing our broader strategy: enabling enterprises to harness AI in practical, production-ready ways that deliver immediate value. We see Cerebrium as a critical enabler for the next generation of AI-powered enterprise applications, from customer service to training, telehealth, media production, and beyond.

Cerebrium co-founder Michael Louis & Maxitech's Investing Partner, Baran Atmanoglu at NYC Tech Week, 2025.

Experienced Founders, Exceptional Results

At Maxitech, we invest first and foremost in exceptional founders who have demonstrated the vision, technical depth, and executional grit to solve hard problems. Cerebrium’s team stands out on all counts.

Michael Louis, Cerebrium’s co-founder and CEO, previously co-founded OneCart, which was acquired by Walmart in 2021. His experience in building and scaling mission-critical systems, combined with co-founder Jono Irwin’s engineering leadership from their OneCart days, has enabled them to deliver remarkable traction in under six months as a lean, focused team.

Their relentless focus on developer experience, performance optimization, and practical infrastructure challenges is evident in the product’s real-world adoption. That level of execution with such a small team speaks volumes about their capability, vision, and understanding of customer needs—a combination we look for in every founder we back.

Cerebrium’s Technical Edge

Cerebrium isn’t simply “another cloud service.” It is purpose-built for modern AI workloads, with features that directly address the toughest challenges in deploying voice and video AI:

• Dynamic Auto-Scaling: Automatically spins up hundreds of GPU-powered containers on demand to handle traffic spikes with minimal engineering overhead.

• Ultra-Low Latency: Platform overhead under 35 ms per request, with co-located pipelines (speech-to-text, LLMs, text-to-speech) eliminating costly network hops.

• Optimized Cold Starts: Container images and orchestration designed for blazing-fast cold starts, mitigating latency spikes during scale-up.

• Hardware Flexibility: Access to over a dozen GPU types, including NVIDIA A100, H100, and L40s, with custom kernel optimizations for large models.

• Integrated Partnerships: Out-of-the-box support for industry-leading ASR and TTS providers like Deepgram and Rime Labs.

• Global Deployment & Compliance: Region-specific deployment for GDPR and HIPAA compliance, with data residency controls.

• Cost Efficiency: Pay-as-you-go billing that charges only for inference time used, with teams reporting up to 40% lower costs than generic cloud providers.

• Optimized Vision & Video Pipelines: Tailored support for computer vision and real-time video workloads — handling frame extraction, batching, and video model inference efficiently for use cases like live stream analysis, object detection, and generative video.

These capabilities make Cerebrium the clear choice for developers who need to deploy high-performance voice and video AI features without investing months (or millions) in infrastructure engineering.

Our Vision for the Future

Maxitech and Cerebrium share a vision of making advanced AI capabilities accessible, reliable, and production-ready for enterprises of all sizes. By abstracting away the complexity of AI deployment, Cerebrium empowers developers to focus on building features that delight users and solve real business problems.

We believe this is the critical infrastructure layer that will enable new categories of enterprise software—from instant voice assistants that rival human conversation to real-time video generation for marketing, training, or customer engagement. Our investment reflects our belief that enabling these use cases at scale will unlock transformative value across industries.

As Cerebrium grows, Maxitech’s role goes beyond capital. We’ll help connect them to our network of enterprise partners, explore pilot programs, and support their go-to-market efforts in North America and beyond. Together, we aim to accelerate the adoption of AI technologies that redefine how businesses operate and how people interact with software.

Ready to Bring Real-Time AI to Your Users?

Learn more about Cerebrium’s technology at https://cerebrium.ai.

Stay tuned for updates on Cerebrium’s journey and Maxitech’s continued investments in transformative enterprise AI.