I’m embarking on a new chapter—migrating my blog from the familiar Hexo setup to a more flexible and intuitive experience blogging framework. This post lays out the upcoming roadmap, the reasons behind the switch, and a peek into the topics I’ve been meaning to write about for ages.

This article describes the final “hygiene” step in an LLM-powered localization pipeline, focusing on programmatically cleaning and validating translated records. It covers removing unnecessary characters, normalizing Unicode, converting punctuation, detecting mismatches and unwanted Chinese characters, adjusting variable spacing, and ensuring all placeholders are present—ensuring translation quality before saving or patching to a TMS.

In the previous installments of this series, we established the foundation for an LLM-powered localization pipeline. In Part 2, we structured our approach further by transforming dictionary data, implementing chunking strategies, and creating translation guides to ensure consistency. We will dive into the critical elements of prompt engineering and validation – the components that determine how effectively our LLMs translate content and how we verify...

In the previous post, we explored how to use LLMs to generate translations in a localization pipeline. Which applied GPT-4o to genererate Full Dictionary Term Base (FDTB). We will continue to build on that foundation and introduce a more structured approach to localization.

Traditional Translation Management Systems (TMS) handle workflow logistics but still rely heavily on the translator’s cultural insight and deep knowledge of both source and target languages. In this blog post, we explore a modern architecture that integrates Large Language Models (LLMs), such as GPT-4o, with a traditional Translation Management System (TMS) to streamline and enhance the localization workflow. The workflow pipeline can ease this dependency...

GPT-4.1 is the latest advancement in OpenAI’s large language models, providing enhanced accuracy and response quality compared to its predecessor, GPT-4o. It builds on the strengths of earlier models while delivering more consistent performance across tasks such as reading comprehension, content generation, and interactive learning tools like quizzes.

SQL Server tracking changes data is useful for auditing purposes, troubleshooting, and data analysis. By tracking changes, you can easily identify what data has been modified, who made the changes, and when the changes were made. Aside with other solutions like CT (Change Tracking), CDC (Change Data Captured), Temporal Table (System-versioning). The most common and simple method for implementing tracking in SQL Server is by using Triggers (Stored...

In this blog post, we will explore the process of deploying a Python Flask application to Azure Web App Service (Linux). However, instead of deploying it to the default wwwroot directory, we will deploy it to a subdirectory within the web app. This approach offers flexibility and organization, allowing you to keep your application code separate from other files in the root directory.