CoreWeave AI Cloud Platform Powers MasterClass’s AI Teaching Agents

CoreWeave, Inc., The Essential Cloud for AI™, announced that MasterClass, the online personalized learning platform, selected CoreWeave Cloud to evaluate and improve the AI teaching agents that power its personalized instruction. MasterClass will use W&B Weave to evaluate how its AI teaching agents perform across conversations with learners, track their behavior, and use those insights to improve agents over time.

AI teaching agents are only as successful as how well they handle an entire conversation with a learner, not just a single question. Evaluating that behavior in production, and improving it continuously, requires both the compute to run agents at scale and the observability to see how they perform with real learners. As AI-native learning tools move from pilots into core product experiences, the ability to trace, monitor, and refine agent behavior in production is becoming as important as the infrastructure running it.

“W&B Weave gives us visibility into complete learner conversations, not just individual prompts, so we can see exactly where our AI teaching agents are succeeding and where they need to improve,” said Mandar Bapaye, chief product officer and chief technology officer at MasterClass. “That level of insight, combined with the performance we’ve seen from CoreWeave Cloud, gives us the confidence to scale.”

MasterClass’s use of CoreWeave and W&B Weave spans evaluation and monitoring, allowing MasterClass to build a dedicated AI loop:

• Full-conversation tracing: W&B Weave captures complete learner interactions with AI teaching agents, giving MasterClass visibility into how agents perform across an entire conversation.
• Production monitoring and custom signals: MasterClass applies custom signals and monitors to live learner interactions to identify where teaching quality and personalization can improve.
• Unified system of record: W&B Weave gives MasterClass a single point of visibility across its evolving AI agent pipeline, from experimentation to production.

Also Read: AiThority Interview with Gou Rao, co-founder and CEO at NeuBird AI

“CoreWeave gives MasterClass a single system of record and the observability their AI teaching agents need to continuously improve,” said Jon Jones, chief revenue officer at CoreWeave. “MasterClass is building on a platform designed to support that full lifecycle, from the compute that runs its agents to the visibility into how well they’re teaching and learning.”

MasterClass is extending this evaluation approach beyond its own engineering team. MasterClass Executive, the first business school experience for the AI era, built with the University of Chicago Booth School of Business and with collaboration from OpenAI, selects a cohort of roughly 700 members each year from a pool of about 30,000 applicants. The includes a hands-on AI lab that gives participants a practical framework for evaluating whether an AI system is ready for production, covering quality, cost, latency tradeoffs, and risk areas such as hallucinations, bias, and inadequate guardrails, using Weights & Biases evaluation and tracing tools.

CoreWeave consistently delivers industry-leading performance, demonstrated by record-breaking MLPerf benchmark results, its position as the only AI cloud to earn the top Platinum ranking in both SemiAnalysis ClusterMAX™ 1.0 and 2.0, and its #1 ranking for inference speed and price-performance for Moonshot AI’s Kimi K2.6 in independent inference benchmarking conducted by Artificial Analysis.