Specialist, QA Test Automation

Cogeco•Burlington, ON

3d•Hybrid

About The Position

We are looking for a sharp, curious Specialist, QA Test Automation who can bridge the gap between strategic vision and day-to-day technical execution. Working closely with the Chapter Lead, QA — who provides the overarching AI QA strategy—you will be embedded within Cogeco’s AI squads to bring that strategy to life. In this role, you will move past standard "pass/fail" scripts to tackle the complex world of non-deterministic outputs. You will focus heavily on probabilistic testing, curating golden datasets, and leveraging cutting-edge LLM evaluation tools like Promptfoo. If you have a solid foundation in QA or software engineering, a healthy obsession with automation, and a restless drive to learn, you’ll fit right in.

Requirements

College diploma or Degree in Computer Science, Engineering, or a related discipline (or equivalent practical experience).
A strong foundation in core QA principles combined with a passion for AI/ML.
3+ years of experience in Quality Assurance, Software Engineering, or Data Engineering.
Experience in Python or JavaScript/TypeScript (essential for configuring evaluation frameworks like Promptfoo and manipulating test data).
Experience with test automation frameworks, API testing, and code repositories (Git).

Nice To Haves

Exposure to Golden Datasets: creating, maintenance, and/or scaling of high-quality golden datasets used as the source of truth for benchmarking model changes.
Exposure to AI Eval Tooling: Champion programmatic evaluation tools such as Promptfoo, DeepEval, or Ragas to automate the assessment of prompt variants and model performance.
Experience working with pipeline tools (e.g., Bitbucket pipelines, Google Cloud run, GitHub Actions, GitLab CI, Jenkins).

Responsibilities

Implement and refine testing workflows tailored for AI powered solutions where outputs are probabilistic rather than deterministic.
Build guardrails to identify model drift, prompt regression, hallucination risks, and safety violations before they hit production.
Own the creation, maintenance, and scaling of high-quality golden datasets used as the source of truth for benchmarking model changes.
Champion programmatic evaluation tools such as Promptfoo, DeepEval, or Ragas to automate the assessment of prompt variants and model performance.
Partner closely with the Chapter Lead, QA to translate high-level AI QA frameworks into actionable, daily tasks for technical teams.
Provide real-time feedback to the Chapter Lead on how the strategy is performing in the trenches, recommending practical pivots based on real-world results.
Collaborate with DevOps specialists to inject AI evaluation metrics directly into modern CI/CD pipelines, ensuring automated gates handle fluid AI responses.
Validate data pipelines feeding our models, ensuring that data integrity is maintained from ingestion to inference.
Analyze data inputs/outputs and debugging prompts to identify the root causes of unexpected AI behavior.
Quickly learn, test, and implement new open-source AI evaluation and observability tooling as the ecosystem evolves.
Thrive in a high-uncertainty environment where standard QA blueprints don’t always apply.
Excel at taking an established roadmap or framework and running with it, troubleshooting technical blockers independently.
Communicate technical testing gaps clearly to both AI developers and product owners.
Be open to constructive criticism, willing to experiment, and value an environment where mistakes are treated as learning milestones.