Senior Software Engineer, Localization & AI (12-Month Contract, Hybrid Vancouver)
My client is a global design and engineering software company. Its tools are used by architects, engineers, manufacturers and media creators around the world. They’re building the next generation of their localization platform, which combines neural machine translation, LLMs and automated quality checks in production. They need an engineer who understands how localization works and also knows how to build, fine-tune and deploy the models behind it.
This is a hands-on role with real influence. You’ll design the AI translation pipeline, set technical standards for localization engineering, and work with localization program leads, product engineering teams and external translation suppliers.
What you’ll work on:
- Designing localization workflows that pair NMT with LLM-based refinement, using frameworks like LangChain
- Building reference-free quality estimation with COMET, BLEURT or custom LLM evaluators, so translation quality can be predicted without reference data
- Building and fine-tuning automated post-editing to improve fluency, terminology accuracy and consistency
- Fine-tuning Hugging Face transformer models on translation memory and glossary data to capture domain terminology and voice
- Building backend services, APIs and integrations that connect product systems, TMS platforms, content repositories and CI/CD pipelines
- Automating content extraction, file processing, translation handoff and linguistic QA for both software and documentation
- Optimizing inference latency, cost and deployment on AWS (SageMaker, Bedrock)
- Benchmarking translation performance with BLEU, chrF++, TER and COMET
- Shaping AI translation strategy, including which editing and post-editing work can be automated
What you bring:
- 4+ years of software engineering experience with a strong backend or AI/ML focus
- Strong Python and hands-on experience with transformer-based models (Hugging Face or similar)
- Experience training, fine-tuning and deploying models on AWS (SageMaker, Bedrock)
- Experience building APIs, microservices and system integrations
- A solid understanding of end-to-end localization workflows: extraction, translation, review, QA and release
- Experience with translation management systems such as Phrase, Passolo or similar
- Experience with SQL/NoSQL databases and cloud-native architecture
- You collaborate well, take other points of view seriously, and can bring engineering, program and vendor teams along with you
Nice to have:
- You’ve built reference-free QE or APE systems
- Experience with LLM orchestration frameworks like LangChain
- Experience integrating localization into CI/CD pipelines
- Experience with automated linguistic QA and validation tools
- A background in NLP, computational linguistics or multilingual data engineering
- Experience optimizing model inference cost and performance in production
The details:
- 12-month contract, T4
- Hybrid in Vancouver, BC: in office one day a week (Wednesdays), with flexibility
- Core hours 9am-5pm PT
- Exceptional candidates elsewhere in Canada will be considered for remote, working Pacific hours
- Interviews are fully virtual: 2-3 rounds covering applied AI, system design and coding
Pay: Up to $75/hr on T4