Senior Agentic AI Developer

Your key responsibilities

• Design, develop, and deploy production-grade AI agents and agentic platforms & workflows for complex business challenges.
• Architect systems that combine agentic harnesses, LLMs, tool calling, external APIs, enterprise data, and existing software applications.
• Build agent orchestration workflows, including state management, context engineering, memory, and human-in-the-loop interactions.
• Define and implement evaluation, testing, observability, and monitoring practices to ensure agent reliability, performance, cost, and security.
• Make architectural decisions that balance autonomy, predictability, scalability, cost, and maintainability.
• Collaborate with clients and multidisciplinary teams to identify opportunities, assess technical feasibility, and translate business needs into practical AI solutions.
• Take ownership of technical delivery, from solution design and prototyping through deployment and continuous improvement.
• Evaluate emerging models, frameworks, and development approaches, and help determine where they bring meaningful value.
• Share knowledge, mentor colleagues, and contribute to the evolution of Osedea's Agentic AI development practices.

The ideal candidate

• 5+ years of professional software architecture and development experience
• Hands-on experience building production-grade agentic systems.
• Strong software development and architecture fundamentals, including system architecture, APIs, asynchronous processing, testing, and cloud deployment.
• Capacity for end-to-end infrastructure design: storage, compute, identity & policy management
• Demonstrated experience designing and delivering AI-powered applications beyond the proof-of-concept stage.
• Solid understanding of agentic architectures, including tool use, orchestration, context management, and retrieval-augmented generation (RAG).
• Experience integrating harnesses, agents or LLMs with external services, internal systems, and existing enterprise data and infrastructure.
• Ability to evaluate AI system quality and reliability, troubleshoot complex behaviours, and make informed technical trade-offs.
• Comfortable working with fast-evolving technologies, tools, standards and best practices.
• Builder mindset with strong architectural credentials and proficiency with agentic engineering building blocks and how they fit together.
• Strong collaboration and communication skills, including the ability to navigate ambiguity, explain technical decisions to clients and teammates, and take initiative.
• Intermediate or conversational French (minimum).
We value proven technical skills, curiosity, and practical experience over a specific academic background or a fixed technology stack.

You should be proficient with

• Programming: Python and modern backend development practices.
• LLM integration: APIs and SDKs for models such as OpenAI, Anthropic Claude, or Google Gemini.
• Agent orchestration frameworks such as Microsoft Agent Framework, LangGraph/LangChain, OpenAI Agents SDK, or equivalent.
• Harnesses & plugin design with platforms such as Codex, Claude code, Pi
• ACP, MCP, Skill standards and other agentic engineering best practices
• Context Engineering methods, including cache hit optimization
• Retrieval methods: RAG, embeddings, vector search, graph search and lookup
• Tool integration: Function calling, API integrations, and patterns for connecting agents to external tools.
• Quality and reliability: Automated testing, agent evaluations, tracing, observability, and error handling.
• Infrastructure: Cloud services (AWS, Azure, or Google Cloud), CI/CD, and production deployment practices.

It's a plus if you have experience with

• Software factories, software development automation
• Model Context Protocol (MCP) and standardized agent-tool integrations.
• Multi-agent architectures and advanced orchestration patterns.
• Agent security, permission management, prompt injection mitigation, and human approval mechanisms.
• Frameworks and tools such as LangSmith, Langfuse, or equivalent evaluation platforms.
• ETL pipelines for ingesting, transforming, and preparing enterprise data for AI applications.
• TypeScript or JavaScript and modern frontend frameworks.
• Enterprise software integrations and workflow automation.
• AI-assisted software development and coding agents.
• Technical consulting or delivering solutions for external clients.

What we offer

Competitive Salary and contribution to your pension plan (RRSP)
Flexible hours of work and choose how you work
Work from anywhere up to 8 weeks
Paid sabbatical
Wellness and productivity spending account
Parental program
Activities
Training
And more...