Carbon nanotube networks have demonstrated promise for enabling printable transistors which can be utilized as switching elements within active matrices. Such active matrices have been utilized for controlling pixels and interrogating sensor elements. A new promising technology for imaging, called an active pixel matrix, aims to integrate both the underlying transistor and sensor element within a single device, namely a phototransistor. Carbon nanotubes are poised to enable printable active pixel matrices for sensing applications, however, due to minimal absorption a secondary material needs to be added as a photo synthesizer. Additionally, the material must be sufficiently electrically coupled with the carbon nanotube network to allow for efficient carrier extraction. Our lab has previously developed techniques for crystallizing P3HT directly onto carbon nanotubes using solution-phase soft epitaxy. In this work, the solution phased technique has been adapted to functional carbon nanotube networks already deposited onto a surface.I will present the performance of a device that consists of a solution-processed carbon nanotube network, which acts as the carrier extraction layer, and a surface crystalized P3HT, which acts as the absorption element. First, I will detail the fabrication scheme and soft-epitaxial method of polymer crystallization. Next, I will overview how the P3HT crystallization affects the electronic performance of the devices, including the photo response. The device exhibits high responsivity, with a maximum responsivity value of 3 x 105. Additionally, the device exhibits a wide spectral response within the visible range, due to the exciton generation at the P3HT/CNT interface. I will also present the transient response of the device, which provides insight into the role of carrier injection/trapped states within the material.Lastly, I will discuss the potential applications for an active matrix of phototransistors, including neuromorphic computing. In these specific applications, the semiconductor can be utilized as the functional material within an Organic Electrochemical Transistor (OECT), where it exhibits behavior associated with mixed ionic/electronic conduction (MEIC). I will benchmark the performance, and discuss various methods to fabricated charge trapped switches, a key memory/computational element for future neuromorphic architectures.