Senior Software Developer, Applied AI

Role overview This remote role in Canada sits within an AI-native engineering pod that builds the internal platform powering an organization's broader software delivery. Rather than shipping product features directly, the senior developer ships the infrastructure—connectors, evaluation harnesses, orchestration tooling, and observability—that allows small engineering teams and non-technical staff across sales, customer experience, finance, people operations, and professional services to use AI safely on real work in a civic and government software context.

Responsibilities - Design and evolve an Agentic SDLC framework, including agent specifications, reusable prompts, orchestration templates, and reusable components for engineering pods. - Build and operate a connector layer: MCP servers and integrations into core systems that respect least-privilege permissions and are treated as versioned, tested, monitored production software. - Maintain evaluation harnesses that automatically score agent output, including regression suites for prompts and workflows and quality gates embedded in continuous integration. - Instrument token spend, latency, eval pass rates, and usage across production agent workflows, and build dashboards and alerts that turn AI cost and quality into a managed system. - Engineer guardrails such as permissioning, audit trails, versioning, and output controls so that agent-assisted work is defensible in a regulated environment. - Ship skill and template libraries, onboarding flows, and self-serve tooling that take non-technical employees from zero to producing real work with AI, with feedback loops feeding back into the platform roadmap.

Requirements - 6+ years as a software engineer shipping production systems, with at least 1 to 2 years building LLM-powered or agentic systems relied on by real users rather than prototypes. - Strong general engineering fundamentals across distributed systems, API design, CI/CD, and cloud infrastructure. - Hands-on depth with current agentic tooling, including frameworks and coding agents such as Claude Code or LangGraph, MCP or comparable tool protocols, structured outputs, and eval-driven development. - A platform-oriented mindset, with success measured by other teams' throughput and a preference for deleting code over defending it. - Comfort operating with a small blast radius and high autonomy within a small pod that carries a company-wide mandate.

Nice to have - Experience in regulated or public-sector software where auditability and defensibility of outputs matter. - Prior ownership of DevOps, platform engineering, or internal developer platforms, including driving adoption rather than just building tools. - Experience instrumenting and optimizing LLM cost and quality at scale, including model routing approaches.