Staff Software Development Engineer - Applied AI

Wagepoint•Calgary, AB

10d•CA$180,000 - CA$200,000•Remote

About The Position

Wagepoint is seeking a talented Staff Software Development Engineer - Applied AI who enjoys tackling hard problems and wants to create meaningful impact for small businesses across Canada. This role involves end-to-end ownership of AI-powered services, from agent architecture through production operation. The position is defined by deep, hands-on ownership of one or more AI services, applying current agentic AI technologies to solve real payroll problems for customers accurately, securely, and at scale. The engineer will be a major contributor to the software factory, the agentic pipeline of specs, agent implementation, agent review, and human approval that produces software. Responsibilities include making sound local architecture calls, selecting agent architectures for cost and predictability, designing data representations that ground LLMs in accurate business context, and owning the full path to production, including application code, infrastructure as code, CI/CD, evaluation, and observability. The role operates within Wagepoint’s established AI platform standards and will be the go-to expert for their domain. Approximately 75% of the time will be dedicated to customer-facing AI products, with the remainder focused on Wagepoint’s internal AI use and the software factory. The company aims for Level 4 on its agentic development ladder, advancing toward Level 5, without compromising quality in a regulated environment. Output is measured by the reliability, cost efficiency, and correctness of shipped products.

Requirements

7+ years of professional software engineering experience, with a proven ability to design, build, and deploy production-grade systems.
1.5+ years building and operating LLM-based or agentic systems in production SaaS or enterprise applications.
Track record of end-to-end service ownership, including application code, infrastructure as code (Terraform), CI/CD, observability, and production support.
Fluency in the current agentic toolchain, including stateful agent orchestration (LangGraph or equivalent), MCP for tool and data integration, and eval/observability platforms (Arize Phoenix, LangSmith, LangFuse, Braintrust, or similar).
Experience designing RAG and retrieval systems on vector-enabled data stores (PostgreSQL/pgvector or dedicated vector databases), and justifying the choice with cost and architecture tradeoffs.
Experience with model selection and routing across frontier and smaller models.
High degree of agency. Comfortable optimizing API performance under real production constraints and building net-new systems from scratch where no established pattern exists yet.
Proficiency in Python with strong engineering fundamentals, and working knowledge of DDD and Clean Architecture principles applied to build highly performant, well-architected systems.
Experience with cloud-native architectures (Azure preferred), including Functions, AKS, and event-driven design.
Experience with security and compliance in AI systems, including prompt injection mitigation, data handling, least-privilege access, and guardrail design (OWASP LLM Top 10 or equivalent).
Strong communication skills, capable of translating technical AI tradeoffs into business outcomes.
A passionate and key contributor to Wagepoint’s software factory, directing and reviewing agentic development while maintaining and improving our high quality and compliance bar.

Nice To Haves

A considered point of view on where a fine-tuned small language model would beat a frontier API call at Wagepoint is a plus, not a requirement.
.NET/C# experience is a plus, not a requirement.

Responsibilities

Own one or more AI-powered services end to end, from agent architecture selection through production operation.
Build data-representation layers that ground LLMs in accurate, token-efficient business context.
Build automated eval pipelines that gate releases, and monitor drift and regressions in production.
Implement guardrails against prompt injection, tool misuse, and data exposure in owned services.
Own Terraform, CI/CD, observability, and cost engineering for owned services.
Connect AI services to Wagepoint’s products and APIs via MCP and secure APIs.
Direct and review agentic development end to end for owned services, moving them toward Level 4 of the agentic development ladder without lowering the quality or compliance bar.