Back/Engineering/Cursor
AdvancedEngineeringCursor

How to Build a Custom AI Translation Service to Replace Expensive SaaS Tools

Learn how a non-engineer built a custom, high-quality website translation service using an AI-native code editor and GitHub Actions. This workflow replaces expensive localization software and agencies, saving over $140,000 annually while improving translation quality and speed.

From How I AI

How I AI: 3 AI Workflows to Save $140K & Automate Marketing with ElevenLabs' Luke Harries

with Claire Vo

How to Build a Custom AI Translation Service to Replace Expensive SaaS Tools

Tools Used

Cursor

AI-first code editor

Step-by-Step Guide

1

Use an AI Code Editor to Build the Core Logic

Using an AI-native code editor like Cursor, build a small, custom server. This approach allows marketers and other non-engineers to build the first 90% of a functional tool before involving an engineering team for final polish and deployment.

2

Set Up a GitHub Action Trigger

Create a GitHub Action that automatically triggers whenever a developer changes or adds a text string in the website's codebase. This action will initiate the translation process.

3

Create Contextual Prompts for Each Language

For each target language, create a dedicated prompt file. This file should contain brand guidelines, a glossary of specific terms that should not be translated, and stylistic nuances to ensure the translation is context-aware and on-brand.

4

Translate via LLM and Commit Back

The GitHub Action sends the new text string and the corresponding language prompt file to an LLM via your custom server. The LLM returns the translated text, which is then automatically sent back to the codebase as a new commit.

5

Integrate with a CMS (Optional)

For content managed in a CMS, add a 'Translate' button to the interface. Hook this button up to the same custom translation service to provide instant, on-demand translations for marketing content.

Become a 10x PM.
For just $5 / month.

We've made ChatPRD affordable so everyone from engineers to founders to Chief Product Officers can benefit from an AI PM.