The AI team is responsible for extracting life-critical insights from large amounts of real-time streaming sensor data. The insights, which target a broad spectrum of e-health challenges, are delivered by our machine learning models that are analyzing data at scale in our cloud-based distributed computing engine. We develop the insights by taking ownership of the complete development process, starting with on-the-ground data collection, and ending with re-iterative model improvement guided through performance monitoring and end-user feedback. Technology-wise, we work in a tech landscape of Python, TensorFlow/Keras, Apache Spark, Golang, SQL, and Kubernetes as the primary languages and frameworks. At the core of our technology is the ML-model architectures, performing temporal and spatial pattern recognition, which are based on deep CNNs, RNNs as well as classic machine learning algorithms, depending on real-time requirements and the complexity of data. We work on applied end-to-end AI solutions that make a difference by keeping up with the state-of-the-art technologies and frameworks.