In many neuroscience experiments, participant behavior can be modeled using monotonic functions. Here, we present a novel Bayesian algorithm for optimal experimental design, specifically developed for these types of function. The method adaptively selects the experimental condition that maximizes information gain on each trial, using Bayesian techniques to guide this process. Its computational efficiency makes it particularly suitable for real-time ("on the fly") paradigms. The algorithm is implemented in Python to facilitate its adoption by researchers without extensive programming or computational expertise. By allowing the user to specify prior beliefs over the expected model, and offering a non-informative prior option in case the expected trend is unknown, the algorithm integrates domain knowledge into a principled Bayesian framework. We illustrate its performance through numerical simulations, showing faster convergence and improved estimation accuracy compared to traditional fixed or random sampling strategies. Furthermore, we validated the approach in an online behavioral experiment, where the method demonstrated robust empirical performance. This tool offers a flexible and efficient alternative for adaptive experimentation in neuroscience, especially in contexts where time or participant engagement is limited and maximizing trial-by-trial information is crucial.