CoNNear neural-network-based hearing aid algorithm that can compensate hidden hearing loss, a recently discovered hearing loss type that originates for the loss of synapses in the cochlea. This algorithm operates as a medical device and is trained to minimizes the difference between the hearing aid that delivers input to an individualized (hearing impaired) model of the human hearing system and a model of a normal hearing person. This results in a hearing aid that reaches the best possible increase in hearing quality considering the non-linear processing characteristics of the human ear and the loss of synapses. This project aims to bring this technology from TRL3 to TRL5 by improving the training of this algorithm both in time and in hearing loss compensation. The outcomes of this project will be combined in an evaluation package that can be integrated in validation studies of hearing aid manufacturers. This will enhance the uptake of our technology and will strengthen our business and valorization plan.