About the role
We are looking for a Frontier Deployed Engineer Intern to join our Automation CoE and work hand in hand with our Frontier Deployed Engineers (FDEs). This is an internal, team-facing role: the FDEs own the relationship with the business, and you help them design, build, test and run the AI agents and automations they deliver.
You will work with frontier AI tools (Claude, Microsoft Copilot Studio, agentic frameworks), Python, APIs and MCP servers, and the Microsoft Power Platform. You will take on well-scoped build tasks, speed up prototyping, keep deployed solutions healthy, and help turn one-off builds into reusable components the whole team can use.
Key responsibilities
Build alongside the FDEs
- Take on scoped build tasks from FDE engagements: agent components, flows, scripts and integrations.
- Rapidly prototype AI agents and automations using Claude, Copilot Studio, Power Automate Cloud Flows and Power Automate Desktop.
- Write Python scripts and lightweight services that connect systems through APIs, MCP servers and Microsoft 365 / Graph.
- Draft and refine prompts, tool definitions and guardrails for agents.
Test, evaluate and harden
- Build test cases and evaluation sets to check agent output quality, accuracy and edge cases.
- Validate solutions against the requirements the FDEs gathered, and flag gaps before release.
- Help move prototypes to production following CoE standards for security, governance and version control (Git).
Run and support deployed solutions
- Monitor deployed agents and bots, triage failures, and fix issues or escalate to the right FDE.
- Pull usage and performance data (run counts, error rates, time saved) so FDEs can report impact.
Strengthen the FDE toolkit
- Turn repeated builds into reusable templates, connectors, MCP servers and code snippets.
- Write and maintain technical documentation, SOPs and runbooks for the team.
- Research new AI tools and techniques and share short findings or demos with the team.
Qualifications
Required
- Currently pursuing a degree in Computer Science, Software Engineering, Computer Engineering, Data Science, Information Technology or a related technical field.
- Hands-on coding ability in Python (preferred) or JavaScript/TypeScript, C# or Java, shown through projects, coursework or prior internships.
- Has built something with AI tools (Claude, ChatGPT, Copilot, OpenAI/Anthropic APIs), even a side project or hackathon entry.
- Comfortable working with APIs and data formats such as JSON.
- Clear communicator and strong technical writer who can keep teammates updated on progress and blockers.
- Team player who takes direction well, asks good questions and follows through on assigned tasks.
- Self-directed and quick to learn; can pick up a scoped task and deliver it with limited hand-holding.
Preferred
- Experience building AI agents, tool use / function calling or MCP servers.
- Familiarity with prompt engineering and basic evaluation of LLM output.
- Knowledge of Microsoft Power Platform, Copilot Studio or Microsoft 365 / Graph API.
- Basic SQL and database knowledge; comfortable with Excel and data analysis.
- Hackathon, open-source or side-project experience shipping working software.
- Bilingual with English and French is an asset.
Technical skills (nice to have)
- AI: Claude / Anthropic API, Copilot Studio, agent frameworks, MCP
- Automation: Power Automate Cloud Flows, Power Automate Desktop
- Code: Python, Git / version control, APIs
- Data: SQL, Power BI, Excel
- Platform: Microsoft Power Platform
What you will learn
- How frontier AI (LLMs and agents) is deployed inside a real enterprise, including security, governance and change management.
- How Frontier Deployed Engineers scope, build and deliver AI solutions, by working inside their engagements.
- Building, testing and evaluating agentic AI solutions with tools, APIs and MCP.
- The full lifecycle of an AI or automation product: prototype, production, monitoring, maintenance.
- The Microsoft Power Platform and Copilot ecosystem.
Success metrics
- Deliver assigned build, test and support tasks for FDE engagements on time and to CoE standards.
- Contribute meaningfully to at least 2–3 AI agents or automations that reach production.
- Measurably reduce FDE time spent on build, testing or monitoring work.
- Add at least one reusable template, connector or component to the FDE toolkit.
- Leave all work documented, version-controlled and handed over cleanly.
Ideal candidate profile
A hands-on builder who wants to learn how AI gets deployed in the real world. You are curious about the newest AI tools, love writing code that works, and enjoy being a dependable teammate who makes the engineers around you faster. Interest in AI agents, automation and the Power Platform is highly desirable.