A cochlear implant (CI) restores sound perception and speech intelligibility to individuals with profound hearing loss. This device transforms sound waves into electrical pulses that stimulate the auditory nerve via an electrode array implanted into the cochlea. CIs have tremendously improved the quality of life of individuals who in some cases have been deaf for decades. While CI users generally have high speech intelligibility in anechoic listening environments, they have difficulty in understanding speech in real-world environments with reverberation. Thus, a strategy that mitigates the effects of reverberation can potentially improve speech intelligibility in CI users. Current algorithms exist to mitigate noise, but to date there is no algorithm that can mitigate reverberation in real-time in a wide variety of acoustic environments. Our research focuses on developing real-time feasible machine learning algorithms to generate cleaner representations of speech to improve intelligibility in reverberant environments.