To maintain consistent communication in noisy environments, conventional mobile communication systems employ retransmissions and robust error correction mechanisms as a strategy to reduce data packet loss. However, as the channel conditions deteriorate, the frequency of retransmissions grows substantially, resulting in notably increased end-to-end (E2E) delays. We propose integrating Network Coding into cloud-native 5G Systems by utilizing the In-Network Computation (INC) infrastructure for swift, low-latency execution, and adaptable, intelligent positioning. We design a flexible network coding model that can dynamically select the appropriate network coding functionality depending on the 5G network condition and specific application requirements.