Agentic Engineer (AI Software Engineer)

About the role

Dayforce is transforming its Canadian payroll and human capital management solution, Powerpay, by applying cloud-native technologies and artificial intelligence. It is looking for a hands-on AI Software Engineer specializing in agentic AI, who will design and build AI agents, automate complex workflows and integrate Large Language Models (LLMs) into enterprise applications.

Key responsibilities

• Design, develop, test and maintain AI-powered applications and intelligent agents.
• Build agentic workflows that execute multi-step tasks, interact with tools and automate business processes.
• Integrate LLMs with enterprise applications, REST APIs, databases and external services.
• Develop AI solutions with frameworks such as LangChain, LangGraph, Semantic Kernel or equivalents.
• Implement Retrieval-Augmented Generation (RAG) so AI apps can use enterprise knowledge.
• Apply prompt engineering, context management and tool-calling techniques.
• Develop reusable backend services and components, and evaluate AI-generated outputs for accuracy, reliability, performance and security.
• Collaborate with developers, architects, Product Owners and QA engineers, and deploy AI services in cloud-native environments.

Required profile

• Hands-on programming experience with Python, Java, TypeScript or equivalent.
• Experience building backend services, RESTful APIs and enterprise applications.
• Understanding of software engineering principles, OOP and design patterns.
• Knowledge of databases, API integration and distributed systems; familiarity with Git, automated testing and Agile.
• Familiarity with AWS or equivalent cloud platforms.
• Ability to work independently, debug, analyze problems and communicate effectively.

Required skills

• Integrating LLMs such as OpenAI GPT or Claude into applications.
• Agentic AI concepts: tool calling, workflow orchestration, planning and context management.
• Prompt engineering, structured outputs and LLM API integration.
• RAG, embeddings and vector databases.
• Understanding of LLM limitations and hallucinations, and of responsible AI, security and data protection.