Staff AI Engineer

Job description

**Build better healthcare together** Maple’s purpose is to meet the world's healthcare needs. We’re Canada's leading on-demand healthcare platform, connecting patients with Canadian-licensed doctors, nurse practitioners and specialists. Today, we’re proud that over 8 million patients and 7,000 businesses and government partners have access to same-day, proactive and ongoing care. We enable people to take control of their health and push the boundaries of what’s possible every day. Bold minds. Big impact. A career at Maple is about making an impact. We empower bold thinking, surround you with an inspiring team and are driven by a motivating purpose. Here, you’ll do work that changes lives—including your own. ### The role We're looking for a Staff AI Engineer to help our engineering teams build, evaluate, and scale AI features across the product — and set the bar for what great looks like. In this role, you'll design the evaluation, safety, and agent execution patterns that every squad builds to, stay hands-on with reference implementations and evaluation harnesses, and coach engineers across the org so squads can ship AI features safely and confidently. You'll work closely with Product Engineering, Platform Engineering, Product and our Security, Privacy, and Clinical partners — turning a fast-moving, regulated healthcare environment into a place where AI can scale with the quality patients and providers deserve. ### Your responsibilities As a Staff AI Engineer, you will: - Set the guardrails, evaluation strategy, and observability standards that keep AI feature quality, safety, and technical health high across the product - Shape the model and provider strategy, usage patterns, and cost controls that keep AI unit economics healthy as we scale - Define the evaluation strategy, agent execution architecture, and reference patterns that the teams building and maintaining AI features build to - Mentor and coach squads and engineers across the org so they can safely do more of the AI work themselves, against Maple's standards - Partner with Product and Engineering squads on agentic feature development as a senior technical partner and reviewer - Co-define the AI governance model with Security, Privacy, Legal, and Clinical partners — including review criteria, PHI handling expectations, and escalation paths - Translate technical AI choices — cost, clinician efficiency, patient experience, and risk — into clear business outcomes for executive and cross-functional audiences ### What success looks like **In your first 90 days,** you'll complete a landscape assessment of Maple's current AI features, evaluation coverage, cost profile, and governance gaps. You'll publish an initial AI standards and best-practices guide covering prompt and agent design, evaluation expectations, safety and PHI handling, and cost hygiene — and stand up an initial evaluation strategy and reference harness that at least one product squad has adopted for an in-flight feature. By month three, you'll have agreed on the ownership interface with Platform Engineering and built the working relationships across Product, Security, Privacy, and Clinical you'll rely on going forward. **Over your first 12–18 months,** you'll roll out a company-wide agent execution architecture adopted by the majority of squads shipping AI features, stand up production-grade monitoring for quality, safety, cost, and latency with clear SLOs, and deliver measurable AI cost optimization against your v1 baseline. Every customer-facing AI feature will be covered by an approved evaluation harness and monitoring dashboard, and squads will be independently shipping AI features to Maple's standards ### What you bring - 8+ years of software engineering experience, with 2+ years in a Staff applied AI or ML engineering role - Demonstrated ownership of agentic or LLM-powered systems in production — including evaluation, guardrails, and cost and latency tuning at scale - A strong track record designing evaluation strategies for non-deterministic systems — offline evals, online evals, human-in-the-loop, regression harnesses - Deep understanding of agent execution patterns (tool use, planning, memory, orchestration frameworks) and their trade-offs - Experience defining AI quality and monitoring practices — quality metrics, hallucination and safety controls, drift detection, incident response - A builder's mindset — a product engineer with deep empathy for user and customer impact, comfortable getting hands-on with code, prototypes, and reference implementations - Proven ability to influence — mentoring senior engineers, setting standards across squads, and partnering with platform teams - Excellent written and verbal communication; you can translate confidently between executives, product, clinical and compliance stakeholders, and deep technical teams **Play to your strengths.** Finding the right fit goes both ways—so this section is yours to complete. When you join, you'll add the unique strengths you bring that aren't captured above. ### Our tech stack We value strong architectural fundamentals and systems thinking over knowing every single tool in our stack, but here is what you will be working with: **AI** - **Frameworks and agents:** LangChain, LangGraph, CopilotKit (for Generative UI patterns) - **Models and foundation services:** AWS Bedrock, OpenAI / Anthropic APIs - **Observability, tracing and evals:** LangSmith and Langfuse for production telemetry, performance evaluation, and prompt iteration - **Data pipelines for AI:** ClickHouse and PostgreSQL powering evaluation datasets and analytics pipelines tied to LangSmith **Platform and infrastructure** - **Backend platform:** Laravel (PHP) core platform integrated with specialized AI microservices and pipelines - **Databases & cache:** MySQL 8, Amazon DocumentDB, Typesense (search indexing), and Redis - **Cloud & DevOps:** Fully containerized, auto-scaling AWS infrastructure defined as code via Terraform and orchestrated through GitHub Actions - **Observability and reliability:** Datadog, BugSnag, and comprehensive testing across Jest, PHPUnit, WebDriverIO, and axe-core ### How we'll support you We recognise our people's health is everything. That's why we take care of them. - **Maple for you and your family:** 24/7 access to general practitioners, pediatrics and mental health therapy for you and your family. Care starts on day one - **Comprehensive health coverage:** Medical, dental and life insurance because your peace of mind—and your loved ones—matter most - **Health spending account:** Extra funds to cover the essentials that make a difference, from new glasses to specialised therapy - **Flex benefits:** An annual budget built for your growth and well-being. Use it for professional development, wellness expenses or a one-time boost to your retirement savings. You choose what matters most - **Health days:** Life happens. We provide 10 dedicated days for rest, medical appointments or caregiving, so you can show up at your best - **Destination days:** Work from anywhere. Enjoy the flexibility to work internationally in eligible countries for up to five days per year - **Group retirement savings plan:** We're here for the long haul. Invest in your future with our group retirement plan ### The details - **Job type:** New role, full-time - **Hiring manager:** VP of Engineering - **Location:** Hybrid, 500 King Street West, 2nd Floor, Toronto, ON, M5V 1L9, OR remote within Canada - **Start date:** October 2026 - **Vacation:** 4 weeks Offers can vary depending on skills, experience and readiness to meet Maple's expectations for this level. We encourage open conversations about pay. If you have questions about how compensation works at Maple, you're welcome to ask at any point in the interview process. We're committed to the safety and security of our platform. Please note that any offer of employment is subject to a criminal record check (EPIC) and identity verification. Additional checks, including employment or education verification, may be conducted depending on the role's requirements. **Use of artificial intelligence** We don't currently use artificial intelligence (AI) or automated tools to screen, assess or select candidates. Every application is thoughtfully reviewed by our Talent Acquisition team or hiring managers.