Back to the Conference Program

Contents

Plenty of attention is being given to the translation aspect of LLMs, but there is less attention given to improving the processes around translation using this technology, even though the major LLM provider technology announcements in 2026 were mostly around software development, rather than language quality. Can LLMs design meaningful localization infrastructure and workflows? Can vibe coding be used to build connectors and file filters? Is localization engineering the branch of localization that can really benefit from agentic workflows? Are translation technology MCPs really useful or just a marketing hype? Is there still a need for DTP and pre-processing? How to transform localization engineering? How to incorporate it if you have never done it before? What platforms to use to host such tools? What are the best practices to avoid overspending and making the most value out of LLMs? This is not a theoretical conversation: the presenter will showcase various process improvements already in production and conversation starter prototypes that may change the way you think. Come prepared with questions and problem statements!

Takeaways

Vibe coding can efficiently solve many localization problems. If junior engineers and PMs skip the learning curve because LLMs handle it, the pipeline of senior talent thins out. PDF text extraction is solved, PDF conversion isn’t. Agentic workflows already can do as good a job on file filter preparation as localization engineers.