About StoryPolish
Bridging the gap between raw translations and immersive reading experiences.
Our Mission
StoryPolish was born from a love for web novels and the desire to bridge the gap between raw translations and truly immersive reading experiences. We believe language barriers shouldn't prevent fans from enjoying stories in their best possible form.
We are dedicated to leveraging the power of cutting-edge AI, specifically fine-tuned large language models like Google's Gemini, to provide translators and readers with tools that enhance clarity, flow, and style. Our goal is to make translated web novels more accessible, readable, and enjoyable for a global audience.
The Technology
We utilize advanced natural language processing (NLP) and fine-tuned AI models trained on vast datasets, including specialized corpora relevant to web novel translations (like parallel translations and well-written fiction).
This allows our system to understand context, grammar, syntax, and stylistic elements specific to the genre, going beyond simple machine translation to capture the nuances that make a story truly engaging.
Meet the Founder

Geoffrey Stephens
Like many, my journey into web novels began with anime – specifically, Dragon Ball Z on Sunday mornings. Decades later, I transitioned to light novels, but grew frustrated with incomplete stories. This led me to web novels, seeking completed Japanese works, only to find the translations often lacked polish.
Switching to Korean and Chinese web novels offered more completed stories, but the translation quality barrier remained. While reading a popular Korean novel one day, inspiration struck: "I can fix this!" I developed a program to refine the text, fed it an EPUB, and was thrilled when it produced over 2000 polished, readable chapters.
I realized others must share this desire for refined translations, and translators likely want their work appreciated by a wider audience. StoryPolish was born from this intersection – a tool to elevate translated web novels for readers and creators alike.