Today, OctoML announced the completion of its $85 million Series C investment round, raising the total funds raised by the business to $132 million. Along with previous investors Addition, Madrona Venture Group, and Amplify Partners, Tiger Global Management led the investment.
“Significant human work is required to increase and fine-tune models before deployment due to the constantly expanding ecosystem of ML hardware backends and various models. This is causing production times to be significantly delayed; expenses to increase, and the inability to implement new use cases in edge devices with limited resources “explained OctoML CEO Luis Ceze. In addition to limiting AI’s accessibility, these issues are leading machine learning down an environmentally unsustainable path.
OctoML is a learning algorithm deployment platform that gives businesses to move their certified models into production choice, automation, and performance. It is built on the Apache TVM open-source framework. The platform aids machine learning (ML) specialists in achieving optimal performance across a variety of cloud services and edge hardware endpoints, leading to considerable cost reductions and expedited time to market. A new generation of smart apps now has more opportunities thanks to the platform’s ability to easily handle the problem of optimizing ML models to fit the resources at the edge.
Using OctoML to put models into production has already resulted in outsized gains for a number of Global 100 companies. These customers experience a 2-10x increase in performance and a 2-10x decrease in computing costs.
The additional funding will support the business’s next phase of expansion, which will include extending its top ML deployment infrastructure and leveraging its expanding ecosystem of hardware suppliers and cloud service vendors.
“According to John Curtius, Partner at Tiger Global, OctoML is causing a significant change in how firms create next-generation AI models and applications.” With OctoML’s goal of providing users with a uniform deployment lifecycle across all the ML hardware vendors they rely on, ML development is becoming more affordable and open to a wider range of developers. We’re thrilled to have Luis and the OctoML co-founding team join the Tiger portfolio and are looking forward to helping them in their next phase of development.