22 Nov
|
Interac
|
Winnipeg
Who We Are
Who We Are: Every transaction matters. Every Canadian matters. At Interac, we protect both — driving trust, security, and inclusion, so our digital economy thrives. Founded in 1984, Interac connects Canadians through secure digital payments, advanced identity verification and industry-leading fraud protection. Connecting banks, businesses, and individuals, Interac enables millions to send, receive, and manage money safely and effortlessly every day — across both digital and physical environments. As the backbone of Canada’s financial ecosystem, Interac facilitates over 20 million transactions daily, supported by trusted partnerships with government and financial institutions. Consistently ranked as Canada’s most reputable financial technology brand, Interac is deeply embedded in the daily lives of Canadians. Who You Will Work With
The Senior Agentic AI Developer reports to the Leader, Data & AI Platforms and designs, develops, deploys, and supports production-grade AI agents and the shared technical harness components that make them secure, reliable, reusable, and governable. The role engineers and maintains agent orchestration, Agent Development Kit patterns, context and memory configuration, MCP server and tool integrations, identity and permissions, human-in-the-loop controls, evaluation, observability, and production tuning across AWS, Microsoft, and Databricks capabilities. The role also works directly with business, product, risk, security, and operations stakeholders to refine requirements, shape agent design, and coordinate testing in a regulated payments workplace. What You Will Do
Design, develop, test, and deploy production-grade AI agents, agentic workflows, and LLM-powered services across Microsoft, AWS, and Databricks that turn business needs into reliable, reusable,
and secure solutions. Engineer and maintain agent harnesses and Agent Development Kit patterns. This includes orchestration, tool execution, memory, identity and permissions, human-in-the-loop checkpoints, error handling, and evaluation hooks, using capabilities such as AgentCore Runtime, Memory, and Identity and Copilot Studio actions and agent flows. Apply context engineering and retrieval-augmented generation (RAG) to control what agents receive and when, including instructions, tool definitions, retrieved knowledge, memory, and state. The aim is to improve accuracy, reduce hallucinations, and manage token cost. Build, secure, and register MCP servers, tools, APIs, and connectors that expose enterprise systems, data, and knowledge sources through approved AI gateways such as AgentCore Gateway. Partner with business, risk, security, governance, architecture, and operations stakeholders to refine workflows, requirements, and success measures. Translate them into agent behaviours, user stories, acceptance criteria, and test scenarios, and facilitate design and testing discussions. Contribute to solution architecture and design, including agent workflows, API contracts, data flows, security controls, and evaluation strategies. Maintain technical documentation and operational runbooks. Apply secure engineering practices, including clean and testable code, automated testing, CI/CD, infrastructure as code,
evaluation frameworks, prompt and model versioning, observability, and content safety, privacy, and governance guardrails. Monitor and troubleshoot production agents, and continuously tune prompts, context, retrieval, latency, quality, reliability, and cost. Help define reusable standards, patterns, and reference implementations for harness engineering, context engineering, MCP tool design, and AI gateway onboarding. Provide technical guidance to developers, citizen builders, and delivery teams. What You Bring
University diploma, certificate, or degree in Computer Science, Software Engineering, Information Systems, Data Science, Artificial Intelligence, or Machine Learning, or an equivalent combination of education and experience. 5+ years of software development experience, including hands-on delivery of production-grade Generative AI or LLM-powered applications. Experience in regulated industries such as fintech, banking, or payments is preferred. Strong proficiency in Python and working knowledge of SQL. TypeScript/JavaScript experience (for example, for MCP servers or Microsoft 365 Agents SDK extensions) and Power Platform experience are assets. Experience with LLM and agent patterns: prompt and context engineering, RAG, embeddings and vector databases, tool calling, MCP, multi-step orchestration, and evaluation. You also understand harness design, including orchestration loops, memory, session isolation, human-in-the-loop controls, and failure recovery. Hands-on experience with Amazon Bedrock, Amazon Bedrock AgentCore (Runtime, Gateway, Memory, Identity, Observability), and Microsoft Copilot Studio, with working knowledge of Databricks. Experience with Strands Agents or LangGraph
📌 Senior Agentic AI Developer (Winnipeg)
🏢 Interac
📍 Winnipeg