SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Trulioo, a global leader in digital identity verification, is seeking a Staff Engineer or Architect to design, build, and own AI agent systems end-to-end for identity, fraud, risk, and commerce workflows. Reporting to the Director of Software Engineering, you will be responsible for creating reliable, steerable, multi-step agent workflows that plan, execute, observe outcomes, and iteratively improve.
Key responsibilities include:
- Design and own agent systems from design through production, including eval harnesses and observability for all shipped features, not just happy-path scenarios
- Build reliable, steerable workflows with tool use and human-in-the-loop patterns appropriate for regulated, high-stakes decisions
- Troubleshoot novel, non-deterministic failures across models, tools, orchestration, and infrastructure, staying calm under ambiguity and finding root causes
- Work within a compliance-heavy, globally regulated domain where auditability, correctness, and human oversight are critical
- Set and communicate technical vision to teammates and direct agents, making implicit requirements explicit
- Leverage AI and emerging technologies to improve productivity and identify continuous improvement opportunities while ensuring responsible, secure, and compliant use
- Orchestrate AI agents in daily work, maintaining a clear mental model of the system and staying accountable
Trulioo is headquartered in Vancouver with strategic hubs in San Diego and Dublin. The role offers a hybrid working model with staff typically working three days per week at the office. The company is recognized as a BC Top Employer and provides comprehensive benefits including health, dental, vision coverage, retirement plans with company match, paid time off, parental leave, and an annual education & training stipend.
Requirements:
- 8+ years of software engineering experience with demonstrated impact on production systems
- Fluency using AI agents in your own engineering workflow; ability to demonstrate how you maintain a mental model, verify outputs, and stay accountable while moving faster with AI
- Demonstrated ability to own features end-to-end and communicate trade-offs clearly to both technical and non-technical stakeholders
- Experience working on challenging, novel, or ambiguous projects where you defined the problem, not just solved a handed-down spec
- Hands-on experience building with LLMs including prompt engineering, RAG, tool use, and at least one agent framework (e.g., LangGraph, CrewAI, AutoGen, or similar), plus evaluation frameworks for agent behavior
Nice to have:
- Experience in financial services, identity, fraud, or risk domains, or other compliance-heavy environments (AML, fraud rules engines, identity graphs, risk scoring)
- Multi-agent orchestration patterns and production-grade evaluation harnesses
- AI observability tooling; MLOps practices (experiment tracking, model registries, feature stores)
- Vector databases, knowledge graphs, or structured retrieval for agent memory
