AgileRL has raised a seed financing round backed by Fusion Fund and Flying Fish, with participation from existing investors Octopus Ventures and Counterview Capital, bringing the company’s total funding raised to $7.5 million.
The company said it will use the new capital to launch Arena in the United States and open an office in San Francisco. Arena is positioned as an RLOps platform designed to make reinforcement learning more accessible for teams building advanced AI systems, including reasoning models and autonomous agents.
AgileRL is pitching Arena as a development and operations layer that abstracts much of the infrastructure burden typically associated with reinforcement learning, which it characterized as complex, expensive, and slow to deploy. The platform includes capabilities for environment, reward function, and dataset validation, along with support for model-free reinforcement learning and fine-tuning large language models. AgileRL also highlighted automated distributed training, hyperparameter optimization, and simplified deployment and monitoring.
To drive early adoption alongside the U.S. launch, AgileRL said it is offering free credits for users who sign up to start training on Arena.
