Microwave radiometers are widely used for the retrieval of liquid water path (LWP) and integrated water vapor (IWV) in the context of cloud and precipitation studies. This paper presents a new site-independent retrieval algorithm for LWP and IWV, relying on a single-frequency 89 GHz ground-based radiometer. A statistical approach is used based on a neural network, which is trained and tested on a synthetic dataset constructed from radiosonde profiles worldwide. In addition to 89 GHz brightness temperature, the input features include surface measurements of temperature, pressure, and humidity, as well as geographical information and, when available, estimates of IWV and LWP from reanalysis data. An analysis of the algorithm is presented to assess its accuracy, the impact of the various input features, its sensitivity to radiometer calibration, and its stability across geographical locations. While 89 GHz brightness temperature is crucial to LWP retrieval, it only moderately contributes to IWV estimation, which is more constrained by the additional input features. The algorithm is shown to be quite robust, although its accuracy is inevitably lower than that obtained with state-of-the-art multi-channel radiometers, with a relative error of 18% for LWP (in cloudy cases with LWP > 30 g m(-2)) and 6.5% for IWV. The highest accuracy is obtained in midlatitude environments with a moderately moist climate, which are more represented in the training dataset. The new method is then implemented and evaluated on real data that were collected during a field deployment in Switzerland and during the ICE-POP 2018 campaign in South Korea.