About

DeconvoluteAI takes its name from a deconvolution (also called transposed convolution) in convolutional neural networks, where this layer is used for upsampling. One application is super-resolution to increase the resolution of images, for which a model learns to fill in the missing information.

I've created this website to share my learnings, document projects, and provide tutorials, in short, to provide information on software engineering and AI.

About me

I am a software development engineer fascinated by machine learning and the human brain. Driven by a passion for understanding systems from first principles, I believe that robust engineering practices are essential for unlocking the potential of machine learning in solving real-world problems.

Previously, I’ve co-created NeuralCeption, a website dedicated to computer vision and machine learning.

I enjoy working across the stack with different technologies, from building products to exploring the latest research in computer vision and natural language processing.