Senior Agentic Developer

Role

Senior Agentic Developer responsible for designing and building agentic systems that enable autonomous agents to plan, execute, self-validate, and iterate. You’ll develop both application logic and cloud infrastructure to support scalable agent operations, create agent skills and families, and ensure reliability in production through tooling, evaluations, and guardrails. This role supports a key client (Audi) and operates in a hybrid model for Toronto-area candidates and remote across Canada on EST hours.

Key Responsibilities

• Develop agents that generate high-quality native applications and production-grade code based on user and business needs.
• Build and maintain automated evaluation pipelines (evals) for agent and skill outputs, including LLM-as-judge scoring, regression test suites, and golden dataset validation.
• Define strict input/output contracts for MCP tools and agent skills using typed schemas; handle edge cases and surface structured errors.
• Own the prompt engineering lifecycle: version-controlled prompt templates, parametric input injection, and structured system/user role separation.
• Contribute to the team’s MCP tooling catalog: implement, test, and document MCP-compatible integrations.

Requirements

• 5+ years of engineering experience with lead ownership and responsibilities in an agile environment.
• Experience with AI coding agents (e.g., GitHub Copilot).
• Experience with CI/CD practices, including GitHub Actions, and strong Git-based version control skills.
• Ability to translate business requirements into clear technical plans.
• Strong teamwork, communication, and mentoring skills.
• Familiarity with accessibility standards (WCAG 2.2) is desirable.

AI Skills

• Design and build agentic workflows and multi-step tool chains or orchestration frameworks with deterministic routing and graceful failure handling.
• Hands-on prompt engineering: structured prompts, few-shot examples, schema enforcement, and output format constraints.
• Experience with RAG (Retrieval-Augmented Generation), function calling, and deterministic routing within LLM-powered systems.
• LLM structured output patterns and evaluation approaches: golden test sets, LLM-as-judge pipelines, and prompt regression testing.

Nice to Have

• Experience defining and rolling out engineering standards (coding conventions, PR workflows, testing mandates, API contracts).
• Experience driving AI tool adoption and creating onboarding/guidance for teams.
• Familiarity with prompt caching, semantic routing, output memoization, and instrumentation of LLM calls (traces, latency, token counts) using tools like LangSmith or OpenTelemetry.
• Familiarity connecting applications to back-end services via RESTful APIs.

Location & Work Arrangement

• Toronto-based HQ in the Distillery District; local candidates in the GTHA are asked to be on-site two days per week.
• Open to remote candidates across Canada; remote hires must work EST hours.

Benefits

• Health & dental benefits and Employee Assistance Program.
• Additional wellness/health stipend and RRSP with matching.
• Extra paid time off perks (birthday off, extra summer holiday day, week-long end-of-year break).
• Hybrid, dog-friendly office with snacks and an active social culture (events, outings, Lunch n’ Learns).
• Access to kyu collective resources, training, conferences, and development opportunities.