A large number of generic AI platforms are available in the market to support training a ML model, use the model for prediction and also label data. In addition, they support ML workflows in the deployment, monitoring and management of models. However, these general-purpose platforms require significant tailoring effort to adequately address domain specific issues. ENTA is developing a platform based on Open Source Kubeflow (kibeflow.org) that supports handling of large-scale network data. Designed with extensive network traffic specific feature extraction capability, it will facilitate quick ML/DL model development. Feature extraction, including custom-defined features, on live interface will be supported. Other standard platform capabilities (e.g., pipeline creation, hyperparameter optimization, training, prediction) will be supported. The platform will be ideally suited for experimentation with various network traffic analytics.