We are a diverse, multilingual engineering group building the future of intelligent assistants. We are looking for a Polish native-speaker engineer who combines strong software skills with sharp linguistic and cultural judgment for their region, and who is energized by using LLMs and AI-assisted workflows to work fast. In this role you will deliver high-quality Polish language-engineering work as part of a larger, fast-moving program - writing code and prompts, generating and auditing data, running evaluations, analyzing the assistant's behavior, and reviewing peers' contributions.

What This Role Is Like:

This is a hands-on, execution-focused engineering role inside a large. You will receive well-scoped tasks from the team and deliver excellent work on them. Because each task is one piece of a much bigger effort, you will not always have visibility into the full end-to-end picture or the reasoning behind every assignment, and access to systems is provisioned on a need-to-know basis. The engineers who are happiest here take genuine pride in craft and precision, work well within clear task boundaries, adapt easily to shifting priorities, and are comfortable making solid progress without needing the complete backstory. If you thrive on focused, high-quality execution and trust in the process, you will fit right in.

Role Impact:

- AI / agentic tooling: continuously look for ways to automate and scale the work below - the natural-language engineering, evaluation, and speech tasks - by building agentic,



LLM-assisted workflows and turning repeated manual steps into reusable tooling.
- Natural-language engineering: improve how the assistant understands and responds in Polish - building and auditing language test data, parse templates, and targeted overrides; fixing interpretation errors.
- Evaluation and failure analysis: run locale evaluations, analyze failure patterns, and write clear native-speaker analyses and quality reports.
- Speech: contribute to speech-recognition and text-to-speech quality - pronunciation, lexicon, and text-normalization work (phonetic transcription a plus).
- Language features: improve handling of dates, times, locations, and event extraction from natural text.
- Review and support: provide a technical check on contributions from data specialists.
- Test-set creation and review: build and review evaluation and test sets that genuinely represent the region and language - reflecting local usage, culture, idiom, and real-world edge cases, rather than literal translations from English.

Key Qualifications

- Native Polish, with deep cultural knowledge of the target region (local usage, idiom, conventions, and how people there actually phrase things).
- Fluent professional English.
- Solid software engineering, especially Python scripting and data manipulation; comfortable with Git and the command line.
- Genuine software-engineering ability - you can read, write, and debug code yourself, not just prompt an LLM to produce it.
- You use AI tools to move fast.

#J-18808-Ljbffr

📌 Polish Language Engineer (Barcelona)
🏢 TELUS Digital AI Community
📍 Barcelona