Senior AI Software Engineer

Stikeman Elliott is building its own agentic AI platform, with production agents working alongside lawyers on the firm’s matters. Built on the Claude Agent SDK and integrated with the firm’s systems of record through MCP tool servers, the platform is already in daily use and continues to expand. This is the senior engineering role responsible for advancing its development.

You will design, build and deploy the firm’s AI applications end to end, including production agentic systems, MCP servers that connect to document, email, matter, billing and CRM systems, and reusable plugins and skills that standardize legal and business workflows.

This is a hands-on senior engineering role that spans application architecture, software development and production delivery. You will also provide technical leadership to a small team, with direct access to the firm’s CIO and Enterprise Architect.

What you can expect:

The opportunity to build greenfield systems using modern tools and technologies, work with a user base of lawyers who actively adopt solutions that deliver value and have the autonomy to shape how the platform is designed and built.

Key Responsibilities:

• Own application architecture, including service decomposition, data persistence, state and concurrency models, failure handling, and component interfaces.
• Build agent-facing capabilities as portable tool servers.

Agentic AI Development

• Build production applications using the Claude Agent SDK, including sub-agent delegation, skills, plugins, memory, scheduled and headless execution, and context engineering.
• Develop and maintain MCP servers and tool sets, including structured outputs, transport, and OAuth.
• Extend reusable plugins and skills that standardize the firm’s legal and business workflows.
• Design human-in-the-loop controls for high-risk agentic actions.

Prototyping, Delivery & Support

• Rapidly translate requirements into working MVPs for user feedback and iteration.
• Take prototypes through to production, ensuring reliability, scalability, observability, documentation, and security.
• Serve as the senior technical escalation point for AI applications.

Integration, Identity & Security

• Build integrations with the firm’s core systems, including document and email management, practice and matter management, time and billing, CRM, enterprise data platforms, and Microsoft Graph.
• Implement secure identity and access patterns, including OBO flows and managed identities, to appropriately constrain agent access.
• Address complex integration requirements, including authentication, session persistence, certificate trust, pagination, rate limits, and idempotency.
• Establish engineering standards across AI repositories, including repository structure, branching, releases, code review, CI gates, environment promotion, and secrets management.
• Lead code reviews for AI applications and plugin contributions, while mentoring developers on engineering practices.
• Build and maintain CI/CD pipelines and strengthen the software supply chain in collaboration with Information Security through automated testing, code quality controls, security scanning, dependency management, and secure application credentials.

Experience Requirements:

• Minimum of five years of experience designing and delivering production software, including at least one year working directly with LLM-powered or agentic applications in production.
• Demonstrated end-to-end ownership of production applications, from architecture and development through deployment and ongoing support.
• Proven ability to take solutions from prototype to production and make sound decisions about technical viability.
• Experience working as a senior technical contributor in a small, hands-on development environment.
• Experience mentoring developers and conducting formal code reviews.

Qualifications:

• Expert proficiency in Python, including asynchronous development, testing, and modern web frameworks such as FastAPI; working knowledge of TypeScript, Node.js, and front-end development.
• Production experience with agentic AI, including modern agent frameworks, tool calling, orchestration, prompt and context engineering, and MCP servers.
• Strong understanding of LLM security and retrieval, including prompt-injection and tool-misuse protections, least-privilege access, data isolation, vector search, embeddings and RAG.
• Experience designing production applications and integrations using REST APIs, OAuth 2.0/OpenID Connect, SQL, vector stores, object storage and distributed caching.
• Strong experience developing cloud applications on Microsoft Azure, including containers, Entra ID, managed identities, Key Vault, managed databases and API Management.
• Advanced knowledge of GitHub and CI/CD, including code review, Actions, environments, secrets, testing and security controls.
• Experience with production observability and AI-assisted development tools such as Claude Code, Cursor, GitHub Copilot or equivalent.
• Strong technical documentation and communication skills

Strong Assets:

• Microsoft Azure, including containers, Entra ID, managed identities, Key Vault, managed databases and API Management. Our platform runs on Azure; if your production experience is on AWS or Google Cloud, we expect you to pick it up quickly.
• Hands-on experience with the Claude Agent SDK, Claude Code plugins and skills, or the Microsoft Graph API.
• Experience building software for lawyers, professional services or other regulated, document-heavy environments.

Salary Range (Toronto Only):

$150,000 - $200,000 annually. The posted salary range reflects the expected compensation for this role. Final offers are determined based on relevant skills, qualifications, and experience.

The position is for an existing vacancy. Artificial intelligence is used to assist in screening and assessing applications for this position.

Stikeman Elliott is committed to accommodating people with disabilities as part of our hiring process. If you have special requirements, please advise Human Resources during the recruitment process.