Model-Assisted Coding is the AI skill of the year

Online Workshop — The MAD Challenge | April 23, 16:00 – 18:00 (Prague time)

Vibe Coding for AI Localization:

Exploring Claude Code and Similar Tools

Model-Assisted Development (MAD) has taken over the world. All development teams are switching to designing specifications and having the code for them generated by Large Language Models (LLMs). Everyone becomes a Builder. How can you apply this in Localization? Explore with us by developing automated translation, dashboards, and orchestrations by chatting with a model.

“Any L10n automation idea is now just one prompt away.”

Or invite your colleagues to join

Session Agenda

Explore AI localization and automation at this workshop, then build customized machine translation tools during the
2-day MAD Hackathon on June 25-26.

Julia Díez López

Software Engineer – Localization at McAfee

Case Study: Localization System for GitHub

Julia will demonstrate building an AI localization system using Claude Code and GitHub. Learn to autonomously localize entire repositories with semantic consistency. She will demonstrate how to connect data to AI models to guardrail multilingual AI responses, drawing on her work with Black Ice.

Jourik Ciesielski

Chief Technology Officer ELAN Languages

Case: LLM Orchestration for Translation

Jourik will demonstrate how to vibe-code TMS connectors and XLIFF-based workflows directly into Crowdin. Instead of using UI elements, he will show how to build specific TMS capabilities that standard tools lack, inspired by Crowdin Dreams.

Having vibe-coded his own TMS middleware from scratch, Jourik brings deep professional expertise to the session.

Breakout Rooms: Team Exercise time

- Try vibe-coding with expert supervision
- Please bring the tools you plan to use for the practical tasks (the focus will be on Claude Code, but any environment is welcome).
- Collaborate with other participants to solve localization challenges in real-time.

The workshop will demonstrate workflows using Claude Code, Replit, and Crowdin, but you are encouraged to use your preferred environments, such as Lovable, Base44, Cursor, and others.

Key Learning Outcomes

Master Claude Code and Next-Gen tools

Learn to set up Claude Code and modern vibe coding environments. Gain the technical foundation used by developers to build high-performance AI localization solutions.

Workflows, Dashboards, and Prototypes

Apply model-assisted logic to AI localization tasks. Build automated workflows and functional prototypes to solve architectural challenges and accelerate your delivery cycles.

Strategic Career Shield

Master AI machine translation automation and vibe coding. Bridge the gap between linguistics and technology to lead in the evolving AI localization market.

Collaborative MAD Integration

Master Model-Assisted Development (MAD) to align with engineering squads. Use their language and tools to lead high-impact AI localization projects within your company’s tech stack.

The MAD Challenge Stages

Start with a free MAD Challenge online in April, or join the deep vibe coding workshop at the GenAI in Localization conference in June.

What to expect:

1h 20m — Masterclass & Demo: Julia and Jourik will show you the logic and tools behind vibe coding in localization.

40m — Practical Mini-Challenge: Apply what you’ve learned immediately in a live, timed session.

Hands-on Practice: We provide the tools and real-world tasks — you build the solution.

Expert Support: Dedicated moderators will be in the breakout rooms to guide you and answer questions.

Result: Leave with your first working automation and a clear path to technical autonomy.

Free Online Session • April 23

What to expect:

Deep Immersion: A 3-day journey from advanced lectures to a full-scale build.

3-Hour Intensive: A dedicated block for creating a robust, professional-grade tool.

Skill Matchmaking: Work with your team or let us match you with a balanced group of engineers and managers.

Industry Showcase: Present your final project to the conference audience and compete for prizes.

How it works:

No advanced coding experience required
Come with your own group (up to 7 people) or work solo
Leave with your first working automation and a clear path to technical autonomy.

What is Claude Code?

Claude Code currently leads the industry in solving real-world programming tasks. For localization, this marks a shift from simply translating strings to building your own automation systems with ease.

Claude Code is an intelligent AI agent designed to work directly within your project. Unlike standard AI chats, it understands your entire folder structure and can execute tasks, fix errors, and build tools in real time.

Full Context: Analyzes your entire AI localization project for total consistency.
Action-Oriented: Executes file updates and processes via vibe coding commands.
Natural Logic: Describe your automation in plain English and let AI handle the execution.

Join MAD Challenge

April 23, 16:00 – 18:00 (Prague time)

Online via Zoom

Bring Your Own Tools

FAQ

What is the main goal of the MAD Challenge?

The goal is to master AI localization through Model-Assisted Development (MAD). Participants will learn how to use vibe coding to build automation workflows and create language-specific tools by interacting with models like Claude Code.

Do I need professional engineering skills to participate?

No. The session is designed for localization managers and linguists. You will use vibe coding—a natural language approach—to build functional AI localization prototypes and automation chains without a traditional programming background.

Will you provide access to Claude Code or other tools?

No, we do not provide software licenses or accounts. Please ensure you have access to the tools you plan to use (such as Claude Code, Replit, ect.) before the session so you can participate in the practical exercise.

How does this workshop help with localization automation?

You will learn to transition from manual tasks to automated precision. The workshop demonstrates how to build custom micro-services and intelligent agents that handle repetitive formatting and LQA, allowing you to scale your AI localization efforts.

What is the difference between the April session and the June event?

The April session is a free online workshop to introduce you to AI localization tools. It serves as a foundation for the MAD Hackathon — a 2-day collaborative competition held during the GenAI in Localization conference on June 25-26.

What is the cost of participation?

Participation in the April workshop and the initial MAD Challenge stages is completely free. This is an open opportunity to explore vibe coding before the main event in June.

GenAI Localization Insider Community

This workshop is part of the Localization Insider Community, a monthly series for localization management professionals.

Each month, localization leaders and project managers gather to analyze strategic frameworks and real-world case studies. Join a peer-learning network to benchmark approaches, test new ideas, and solve shared industry challenges.

Vibe Coding for AI Localization:

Exploring Claude Code and Similar Tools