Inferless is joining Baseten! After launching in 2022 to tackle serverless deployment bottlenecks, we realized that scaling for modern AI applications requires a more comprehensive stack. By joining Baseten, we aim to combine our expertise to better support developers who need to run mission critical AI applications with high performance, reliability and cost efficiency.
The Inferless Journey
When we started Inferless in December 2022, our mission was simple: build a true serverless GPU inference platform that helps developers deploy any model in minutes. This came directly from our own challenges building AI applications. Back then, developers were wrestling with fundamental questions like how do you deploy a model reliably and scale it effectively while maintaining high performance? We decided to be laser-focused on solving production bottlenecks which ultimately led to the birth of Inferless.
We saw something that felt obvious but wasn't widely discussed: the inference problem was going to be massive, complex, and fundamentally different. Of all the challenges in inference infrastructure, cold starts became our obsession. A problem that seems simple on the surface but has massive implications for how AI systems work in production. Along the way, we helped hundreds of developers deploy their production workloads, but we realized that to do inference really well and capture the whole stack for enterprises, we needed more comprehensive tooling.
When Amir and the Baseten team shared their mission, it immediately resonated with us. We have been in the AI infrastructure space for years, watching the rapid improvement of open-weight models change the game entirely. Teams now run portfolios of models—frontier APIs for some tasks, open models for high-volume workloads, fine-tuned models for specialized performance. Making this work in production requires solving genuinely hard infrastructure problems, exactly what Baseten has focused on for over years. We kept asking ourselves: where can our expertise have the most impact? The answer became clear: Baseten.
"Talent is the most important ingredient in the AI infrastructure race and the Inferless team is bringing invaluable expertise to Baseten to help bring improvement to the Baseten Inference Stack. I am honored to welcome this talented group to the team and can’t wait to see the impact they will have for our customers."
What’s next
The opportunity ahead is enormous. AI infrastructure is still in its early innings, and teams are just beginning to figure out what production-grade AI systems require. We're excited to bring our experience and obsession to Baseten's mission of powering the world's AI-driven applications with the most performant, scalable, and reliable infrastructure for machine learning inference. To the Inferless community, investors, and friends: thank you for trusting us and making this journey possible. We are looking forward to working with the amazing team here.
