Junior AI Engineer

Position Overview

Canada Quant Labs is seeking a Junior AI Engineer to join their team in Victoria, British Columbia. This hybrid role offers the chance to engage in hands-on work with Canada's public open-weight AI models under mentorship from senior engineers. The role welcomes candidates early in their careers, including self-taught individuals, bootcamp graduates, career changers, and recent graduates without degree or industry experience requirements. The job is full-time and permanent.

About Canada Quant Labs

Canada Quant Labs specializes in training, quantizing, and deploying sovereign AI models on Canadian Blackwell hardware, releasing open-weight models, recipes, benchmarks, and detailed engineering logs publicly through their Hugging Face organization.

Key Responsibilities

• Develop and maintain quantization evaluation harnesses that test W4A16 and NVFP4 quantized models across reasoning, instruction-following, and long-context benchmarks, producing critical go/no-go decisions for releases.
• Create and manage dataset pipelines to process, clean, and decontaminate Canadian corpora training and evaluation datasets across specialized domains such as legal, medical, and financial sectors.
• Conduct inference benchmarking on vLLM serving performance using DGX B300 systems by reproducing published benchmarks, identifying failure points, and documenting causes.
• Contribute to writing model cards and engineering logs that document measured outcomes; outputs are openly shared with contributors' names acknowledged.

Candidate Profile and Qualifications

• Based in British Columbia and at the start of your career: self-educated, bootcamp graduate, career switcher, or recent graduate with no mandatory degree or previous work experience.
• Proficiency in Python with ability to script efficiently on large log files (e.g., 2 GB) and familiarity with virtual environments.
• Experience with PyTorch is advantageous; projects or academic exposure to transformers or large language models is a plus but not necessary.
• Strong curiosity about open-weight models, quantization techniques, and inference mechanisms; personal experimentation such as running models on hardware is appreciated.
• Availability to work from the Victoria office on a hybrid schedule combining remote and onsite days.

Desirable Extras

• Personal projects that deliver tangible outputs like trained models, benchmarks, or widely used tools.
• Participation in competitions like Kaggle, hackathons, or CTF events demonstrating problem-solving aptitude.
• Open source contributions at any scale showcasing ability to navigate unfamiliar codebases.
• Technical writing such as blogs or detailed project write-ups.

Additional Details

• Compensation ranges from CAD 85,000 to 115,000 plus equity, reflecting experience level.
• This is a hybrid role located in Victoria, BC, requiring physical presence on some days.
• The hiring process includes an introductory call, a technical discussion focused on a candidate’s project, and then an offer.
• Applicants are encouraged to submit a résumé and portfolio links; no cover letter is required but a brief motivation is requested.