Data Scientist, LLM Systems

About the Position

Would you like to become part of the LingQ team? We are always looking for exceptional, passionate people to join our team and help make LingQ better! At the moment, we are looking to fill the following position(s):

Data Scientist, LLM Systems

Location: Remote (Canada)

Type: Full-time

LingQ is a Vancouver-based web and mobile language learning app.

You’ll work on implementing and improving the health, performance, and evolution of our AI lesson preparation systems and AI learning assistant - your work involves annotating data, building test suites, finetuning LLM models, designing and improving systems, and integrating the latest LLM innovations - all with an eye to reducing costs, boosting lesson data quality, and speeding up delivery for millions of learners.

Responsibilities

Test Suite & Metrics

• Design, implement, and maintain a modular, reusable test framework for a variety of AI outputs.
• Define and track KPIs: accuracy, latency, cost per token, failure rate.

Model Research & Optimization

• Benchmark, A/B test, and upgrade LLMs as needed from OpenAI, Anthropic, Google, and HuggingFace.
• Optimize model selection and prompt templates to minimize cost and latency while maximizing quality.

Data Annotation & Quality

• Lead multilingual data-annotation initiatives (Spanish, French, Portuguese, Italian, German, Russian, Korean, Chinese - fluent in at least two).
• Establish annotation guidelines and QA processes to ensure consistency and reliability.

Cross-Functional Collaboration

• Work closely with QA testers, backend engineers (Django), ML engineers, and product managers to gather requirements and roll out improvements.
• Translate business needs into technical specifications and deliverables.
• Track downstream errors (logs, user complaints, strange AI outputs) back upstream and fix at source (model-level).

System Optimization

• Stay abreast of AI/LLM trends, multilingual models, cloud server costs, and major provider roadmaps in order to lower our system cost, improve reliability, improve speed, and output quality.

Requirements

• 2+ years in data science or machine learning, with demonstrable hands-on LLM experience.
• Proficiency in Python and experience with the OpenAI, Anthropic, Google, and HuggingFace APIs.
• Hands-on with finetuning LLMs and evaluating model performance (accuracy, cost, speed).
• Multilingual aptitude: fluent or very comfortable annotating in at least two of {Spanish, French, Portuguese, Italian, German, Russian, Korean, Chinese}.
• Passion for language learning and an appreciation of linguistic nuance.

Attributes We Value

• Detail-Oriented: you catch edge-cases before they reach production.
• Communicative: you clarify requirements, share progress, solicit feedback across functions.
• Creative Problem-Solver: you think beyond standard approaches in order to fix system issues and innovate new solutions as necessary.
• AI enthusiast: you proactively scout and pilot new AI/LLM innovations to improve our AI systems and stay up to date on changes in AI.

Nice to Have

• Familiarity with Django or similar backend frameworks.
• Experience building CI/CD pipelines for ML deployments.
• Knowledge of cloud infrastructure (AWS, GCP) and containerization (Docker, Kubernetes).

We Offer

• Competitive salary.
• Fully remote team with flexible hours.
• The chance to shape the future of AI-powered language learning at scale.

About the Company

LingQ is breaking new ground in the multi-billion dollar language learning industry and catering to the universal desire to explore other languages and cultures. Become part of an exciting and rewarding project. Help us continue building LingQ into the premier language learning app.

How to Apply

Please send your resume and a cover letter explaining why you are the one to join our team to jobs[at]lingq[dot]com.

Even if you don't see a job here that suits your skills, we are always looking for passionate, skilled people looking to help us in our mission.

We thank all applicants for your interest. Unfortunately, time allows us to only contact those candidates who are short-listed.

We look forward to hearing from you!