About Us

Kaiops service is available as an AIaaS (AI as a Service). In our AIaaS architecture, the AI agent acts as an external consultant to the IT network, which requests action recommendations from the AI agent by providing relevant network data. The AI agent uses provided data to update models, infer optimal action, and respond with an action or prediction (e.g., scale resources OR respond to security threat). Kaiops leverages service meshes and serverless architectures (e.g., Istio, Knative) by making use of out-of-the-box cluster observability, security, and eventing, efficient and consistent implementation. Kaiops uses specialized Machine Learning (ML) models that expand and augment traditional Reinforcement Learning architectures, pulling from other areas of Machine Learning, including Autoencoders, Generative Networks, and Transformers. Deep Learning models are implemented with open-source neural network libraries (e.g., TensorFlow, PyTorch).