EXPERT COURSE

AI Translation (NMT+LLMs)

Expert Course with Andrés Romero Arcas

AI has rewritten the translation workflow. Machine translation is no longer an experiment at the edge of localization. Neural engines and large language models now sit at the centre of how multilingual content gets produced, reviewed, and shipped. For translation professionals, the question is no longer whether to work with AI, but how to direct it with judgement, control, and measurable quality.

Knowing how to run an MT engine or prompt an LLM is only the starting point. The professionals who stand out are the ones who can choose the right approach for each scenario, decide when to fine-tune, when to inject terminology, and when an LLM-based or agentic workflow is the better fit - and then prove the result with sound quality evaluation.

From Using Tools to Designing the Strategy

Across five focused sessions, you will move from the current AI translation landscape through NMT and LLM solutions and into both automatic and human quality evaluation, finishing with a practical framework for defining your own AI translation strategy. The goal is real-world execution: building AI-informed workflows that scale without compromising linguistic quality or cultural nuance.

By the end of this course, you will walk away with a clear framework for defining, deploying, and evaluating AI translation strategies that scale without compromising quality.


Conditions: Please read our course and subscription plans terms and conditions carefully. With your registration, you confirm that you have read, understood and accepted our conditions and agree with them. 

If you have any questions, please visit the FAQ section (for courses or subscription plans) or get in touch with us.

Add to Calendar

Apple Google Office 365 Outlook Outlook.com Yahoo

  • This webinar includes:
  •     Expert tutor: Andrés Romero, AI Specialist at Acolad Group
  •  Lifetime access to the course and extra contents
  •  In English
  •  TranslaStars Certificate of Completion
  •  Course duration: 10 h (approximately)
  •  When: 16-20 November 2026 (18.00 CET)
Course created in collaboration with:

Free for Crowdin Enterprise Users

AI Translation: NMT, LLMs

Master AI translation with NMT engines, LLM solutions, prompting,
RAG, fine-tuning and MT quality evaluation

Session Breakdown

Session 1: AI Translation Landscape

Machine Translation overview and context. Fundamentals and why incorporate MT. Adoption challenges. NMT vs LLMs: capabilities, limitations, and use cases. Solutions for current limitations and which ones to implement.

Session 2: NMT Solutions

Fine-tuning (training): definition and purpose. Providers and the importance of data. NMT engine training process and retraining. To train or not to train? Other alternatives: terminology injection and adaptive Machine Translation.

Session 3: LLM Solutions for Translation: Prompting, RAG and Fine-tuning

LLMs: definition and development process. Challenges for adoption and advantages for translation. Prompting. Retrieval Augmented Generation (RAG). Fine-tuning. Agentic solutions. Are LLMs the future? Key points.

Session 4: Machine Translation Quality Evaluation I

Automatic quality evaluation: definition and purpose. Reference-based metrics (lexical and semantic). Reference-free metrics. Key ideas.

Session 5: Machine Translation Quality Evaluation II and AI Translation Strategy

Human evaluation and the role of human linguists. Context and market dynamics. The concept of quality. Deployment, quality monitoring and management. Key ideas and recommendations for defining your AI translation strategy.

Course Description

What Is This Course About

This course gives a structured, end-to-end view of modern AI translation. It begins with the current machine translation landscape and the choice between NMT and LLM approaches, then explores NMT solutions such as fine-tuning, terminology injection, and adaptive MT, followed by LLM solutions including prompting, Retrieval Augmented Generation (RAG), fine-tuning, and agentic workflows. The final sessions cover machine translation quality evaluation, both automatic metrics and human evaluation, and close with practical recommendations for defining your AI translation strategy.

Topics Covered

▸ The modern AI translation landscape: NMT vs LLMs
▸ NMT solutions: fine-tuning, terminology injection, and adaptive MT
▸ LLM solutions: prompting, RAG, fine-tuning, and agentic approaches
▸ MT quality evaluation: automatic metrics and human evaluation
▸ Building and deploying your AI translation strategy

Who Is This Course For

This course is designed for anyone who shapes, manages, or delivers multilingual content and wants to put AI translation to work with confidence.

It is especially valuable for:

▸ Localization and Project Managers who need to define scalable, high-quality translation strategies and decide where AI fits in their pipelines
▸ Localization Engineers and Solution Architects who design and implement complex MT and LLM workflows and want a clear view of the options and trade-offs
▸ Translators and Linguists who want to understand the technology behind the tools they use every day and grow into higher-value quality and oversight roles
▸ Language Service Providers and teams scaling operations who need consistent quality and smart automation across languages

No engineering background is required. If you work with language and want to lead in the era of AI-assisted translation, this course is for you.

What You Will Learn

Across five focused sessions, you will move from foundational concepts to strategic execution. By the end of this course, you will be able to:

✔ Navigate the AI translation landscape and distinguish between NMT and LLM approaches
✔ Apply NMT fine-tuning, terminology injection, and adaptive MT strategies
✔ Use LLMs for translation through prompting, RAG, fine-tuning, and agentic workflows
✔ Evaluate translation quality with automatic metrics (reference-based and reference-free)
✔ Design human evaluation protocols and integrate them into scalable quality strategies
✔ Define a complete AI translation strategy with monitoring, quality management, and continuous improvement

Requirements & Setup

Prerequisites

No prior engineering or programming knowledge is required. Sessions are explained from the ground up. A basic understanding of translation workflows is recommended.

Technical Setup

▸ A laptop with a stable internet connection
▸ A free account with a major LLM provider (ChatGPT, Claude, or Gemini) - recommended for hands-on parts
▸ Optional: access to a machine translation provider or platform for live demonstrations

Andrés Romero Arcas

| Language Technology Expert
| Linguistic Engineer
| Machine Translation and AI Specialist at Acolad Group
About ANDRÉS
Andrés is a proficient language technology expert with over a decade of experience in the Localization Industry.
Throughout his career, he has held diverse roles, such as CAT Tool Specialist, Localization Engineer and Operations Technology Coordinator, where he led a team of localization engineers.
Currently at Acolad, Andrés focuses on machine translation evaluation and engine training. He is also deeply involved in prompt engineering and Generative AI, proposing AI-driven driven solutions to deliver tailored, customer-centric solutions and to tackle challenges in Production.
Andrés is passionate about automating and optimizing processes to enhance productivity and efficiency, improving quality and integrating innovation into localization workflows.