In a strategic move to solidify its position as a leader in AI infrastructure, Livingston-based CoreWeave, the “AI Hyperscaler,” announced a definitive agreement to acquire OpenPipe Inc., a platform specializing in training AI agents with reinforcement learning (RL). The acquisition builds on CoreWeave’s recent efforts to vertically integrate its technology stack, following its acquisition of AI developer platform Weights & Biases.
The deal brings together CoreWeave’s high-performance AI cloud with OpenPipe’s advanced self-learning technology. OpenPipe has developed a widely-used open-source RL toolkit for training agents, known as Agent Reinforcement Trainer (ART). This technology enables AI models to learn from experience, continuously increasing in accuracy, performance, and reliability over time. The company’s tools are designed to democratize reinforcement learning for developers and enterprises, bridging the gap between a prototype and a production-ready system.
Brian Venturo, co-founder and chief strategy officer at CoreWeave, emphasized the strategic importance of the acquisition.
“Reinforcement learning is emerging as a pivotal force to strengthen model performance on agentic and reasoning tasks,” he said. “By combining OpenPipe’s advanced self-learning tools with CoreWeave’s high-performance AI cloud, we’re expanding our platform to give developers at AI labs and beyond an important advantage in building scalable intelligent systems.”
OpenPipe CEO Kyle Corbitt echoed this sentiment, noting that the combination with CoreWeave will accelerate the vision of bringing their technology to a wider set of enterprises eager to build more reliable, higher-performing, and cost-effective AI systems.
The acquisition of OpenPipe and the previous acquisition of Weights & Biases highlight CoreWeave’s ongoing strategy to provide a more comprehensive, vertically integrated solution for its customers, offering greater flexibility to train, adapt, and optimize their AI models.
The terms of the acquisition were not disclosed.
