Job Description What''s the opportunity? RBC''s AI Group is building the reusable patterns that help teams across the organization design, build, and ship AI agents faster.

We''re seeking a Staff AI Engineer to design and build these patterns agent identity, context abstraction, and a standardized data and analytics framework that feeds a central evaluation engine for agent reliability.

This includes leveraging agent exhaust data to power an analytics engine built for reuse, discoverability, and value generation across teams.

This hybrid role combines hands-on AI and agent development with full-stack ownership, for someone who wants to give AI builders across the organization the tools to move from idea to shipped agent faster.

Your responsibilities will include: Designing and building agentic and AI-driven logic including LLM-as-judge and rubric-driven prompting patterns that turns raw platform data into decision-ready signals for the business Building and orchestrating agentic workflows designing multi-step agents, tool integrations, and evaluation harnesses that reliably automate complex, governance-sensitive tasks beyond single-shot prompting.

Building full-stack backend services, APIs, and the executive-facing dashboards and reporting flows that put those signals in front of stakeholders - you ship your own AI logic into production, not just a notebook Building and owning the data pipelines that feed your applications, partnering with Data Engineering on underlying data contracts and with Governance/Security on privacy-safe handling of sensitive data Documenting design decisions, prompt/rubric changes, and evaluation results to keep AI-driven components auditable, especially where they touch governance-sensitive workflows.

You''re our ideal candidate if you: Have staff-level experience spanning applied AI/agentic engineering (agent architectures, prompt design, evaluation, LLM-as-judge patterns) and full-stack software development, this is a genuine dual-competency role, not an AI specialist who dabbles in UI work.





Demonstrated track record designing and delivering enterprise-grade AI platforms with direct, hands-on engineering contribution at the most complex levels.

Full-stack capability, hands-on7+ years building and shipping production web applications end-to-end backend services and APIs (e.g., Node.js/Python/Java or equivalent) plus a up-to-date frontend framework (React, Angular, or equivalent) , you can independently take a feature from data contract to a working executive-facing UI without a separate frontend engineer.

Are comfortable owning ambiguous, high-judgment problems deciding what ''good enough'' classification accuracy looks like and when to escalate beyond prompting is a core part of this role Have shipped production systems that consume LLM outputs as a decision signal, not just as a chat interface Can work credibly with Governance/Compliance and Security partners on privacy-sensitive, auditable AI decision-making, given the misuse-detection use case touches individual employee behavior Enjoy working across the stack rather than staying in a single layer, this role spans data contracts, backend, AI logic, and the executive-facing interface, with no expectation of a dedicated frontend or backend specialist to hand off to Tech stack you''ll work with LLM & AI platform: LLM Gateway; LLM API integration and production deployment patterns (Anthropic Claude, OpenAI, Azure OpenAI, or equivalent); prompt/rubric design for LLM-as-judge classification; agentic AI concepts (MCP, A2A, guardrails, authentication, evaluation) AI frameworks: Lang

Chain, Llama

Index, or similar for orchestration/evaluation tooling API & backend: RESTful API design, API gateway patterns, authentication/authorization (OAuth2, JWT), rate limiting,



versioning and deprecation practices Data & retrieval (as needed for classification context):Vector databases / RAG pipeline patterns (Pinecone, pgvector, Chroma, or equivalent) if semantic retrieval is needed to support classification accuracy Cloud & infra: Azure and/or AWS; containerization and cloud-native deployment (Kubernetes/AKS/Open

Shift); Terraform/infrastructure-as-code exposure is a plus.

Full-stack (hands-on, required):Backend service/API development (Node.js, Python, or Java/equivalent) plus a modern frontend framework (React, Angular, or equivalent) for the executive-facing dashboard and Top-10-spenders reporting UI; relational/analytical data store experience (e.g., Postgres or equivalent) for the aggregation layer Data Pipelines & Warehouse: Apache Airflow/Prefect for ETL/ELT orchestration, Snowflake ,Custom Python: Pandas, PySpark for transformations for data lineage and SQL-based transformations What''s in it for you? Become part of a team that thinks progressively and works collaboratively.

We care about seeing each other reach full potential; A comprehensive Total Rewards Program including bonuses and flexible benefits, competitive compensation, commissions, and stock options where applicable; Leaders who support your development through coaching and managing opportunities; Ability to make a difference and lasting impact from a local-to-global scale.

About the AI Group: RBC''s AI Group is the AI accelerator for RBC, with a focus on driving the shift from early-stage AI projects to scaled, client outcomes that amplify the impact of RBC''s people.

In addition to helping scale the biggest AI opportunities at RBC, the AI Group is responsible for advancing research into emerging use cases across generative and agentic AI, while maintaining expertise in security, responsible AI and regulatory expectations.

The Business Enablement function within the AI Group partners with LOBs and Functions to set AI ambition, originate transformation opportunities, and frame programs for delivery ensuring RBC remains at the frontier of AI-enabled value creation.

📌 Staff AI Engineer (Toronto)
🏢 0000050007 Royal Bank of Canada
📍 Toronto