AI quality and security intern evaluating agentic systems at Kinaxis, a supply-chain orchestration software company. Building evaluation, observability, reliability, and security tooling.
Responsibilities
Work directly with a senior architect on the machinery and methodology for evaluating, monitoring, and investigating agentic AI systems
Complete a scoped project within the AI assurance toolkit focused on evaluation infrastructure, quality engineering, security analysis, or tooling for understanding agent behavior
Extend the evaluation harness and improve configuration, execution, reproducibility, and comparability of evaluation runs
Investigate scoring, aggregation, run-to-run variance, sample size, and confidence in evaluation results
Research agent evaluation tooling and techniques and prototype promising ideas
Investigate AI-agent security and trustworthiness, including sensitive context, policy compliance, unsafe instructions, tool errors, and reviewable evidence
Build observability and investigation tools for trace analysis, failure categorization, regression detection, and engineering reports
Use a GitHub-native, agentic engineering workflow with AI-assisted development tools
Present findings and form opinions about the work
Collaborate with experienced software engineers across quality engineering, AI evaluation, reliability analysis, and security-minded investigation
Requirements
Completed 3rd year of, or recently graduated from, a program in Software Engineering, Computer Engineering, Computer Science, or a related field
Currently enrolled in full-time education or a recent/upcoming graduate within 12 months of the placement end date
Demonstrated interest in software quality, reliability, security, or investigation-oriented engineering
Demonstrated interest in AI agents and their evaluation
Comfortable writing code in Python or a comparable language
Comfortable with Git and GitHub
Clear written communication
Comfortable with ambiguity
Nice to have: experience with test automation frameworks, CI/CD pipelines, property-based, fuzz, or metamorphic testing
Nice to have: exposure to application security, threat modeling, policy testing, red-team style evaluation, secure software development, or responsible AI risk assessment
Nice to have: exposure to LLM applications, RAG, prompt engineering, agentic frameworks, and evaluation tooling
Nice to have: familiarity with LLM-as-judge, rubric-based scoring, or benchmark design
Nice to have: experience analyzing logs, traces, audit records, experiment results, or failure reports
Nice to have: experience with data manipulation, analysis, data quality, or validation techniques
Nice to have: exposure to knowledge graphs, semantic modeling, or ontologies
Nice to have: exposure to supply chain, logistics, manufacturing, or enterprise planning systems
Nice to have: contributions to open-source projects, research, competitions, or hackathons
Benefits
Flexible vacation and Kinaxis Days (company-wide days off)
Flexible work options
Physical and mental well-being programs
Regularly scheduled virtual fitness classes
Mentorship programs, training, and career development
Application Security Researcher building autonomous application - security capabilities for OX Security. Researching attack paths, AI models, and detection engines from prototype through production.
Managing Norfolk County’s cybersecurity, disaster recovery, incident response, and security projects. Supervising IT security staff and protecting municipal systems, services, and information.
IT Security Officer overseeing information security for Desjardins’ property and casualty insurance sector. Managing risks, vulnerabilities, suppliers, complex projects, and executive security reporting.
Azure SA&A Specialist supporting Government of Canada security assessments, controls, evidence, and testing. PLATO delivers Indigenous - owned software testing and technology services across Canada.
Senior Security Engineer owning identity, IT automation, and cloud security for Maze, a user research platform. Driving vulnerability management, compliance, and secure infrastructure at scale.
Product Security Engineer securing WorkOS developer tools and APIs for enterprise authentication and identity. Leading secure design, offensive testing, tooling, and vulnerability remediation.
Product Security Lead securing Miovision’s intelligent transportation products, cloud services, and connected devices. Leading secure - by - design engineering and product risk management for safer smart transportation networks.