Magical is building agentic AI for healthcare operations. This is not AI for demos. Our agents have helped suicidal veterans get matched to care faster, identified patients who may have undiagnosed cancer, scheduled patients who otherwise would have waited, and helped providers get paid for the care they deliver. Plainly stated, we impact people's lives. The work is technical, messy, and complex. The impact is real. Our agents operate across real healthcare environments: portals, documents, browser workflows, queues, model calls, human review, edge cases, and exceptions. We are not building toy workflows. We are building systems that need to work reliably in production. Magical is backed by Greylock, Coatue, and Lightspeed, with $41M raised. Our executives have sold businesses to LinkedIn, and have previously sold billions of dollars. .

Why Join as a New Grad

We are hiring up to 5 New Grad Software Engineers to join our Toronto engineering team. This is not a traditional new grad role where you spend a year waiting for real ownership. You will work on real systems, real customers, and real healthcare workflows from the start. You will be expected to learn quickly, ask sharp questions, take ownership, and make progress through ambiguity. You will also not be thrown into the deep end without support. Magical has exceptional senior technical leadership, experienced engineers, and strong mentorship in place. You will get training, feedback, and direct exposure to people who have built and scaled serious products before. We are looking for new grads who want a steep learning curve, meaningful responsibility, and the chance to build AI systems that are already being used in production. This is a hybrid role based in our Toronto office.

About the Role

As a New Grad Software Engineer, you will help build the systems that power Magical’s AI agents. Under the hood, Magical orchestrates long-running, high-stakes healthcare workflows involving queues, run state, browser automation, model calls, evaluations, observability, and human-in-the-loop recovery. These systems need to be reliable, debuggable, and resilient because they operate in messy real-world environments. A workflow might depend on a payer portal changing, a document being malformed, a model output needing review, a human needing to step in, or an edge case no one has seen before. You will help turn messy customer inputs — SOPs, decision trees, payer rules, portal behavior, operational know-how, and edge cases — into reliable automations. You may work on systems that draft workflows, test automations, catch failures, recommend fixes, escalate to humans, improve runtime reliability, or make our agents easier to build and deploy. The problems will not always be well-defined. That is part of the job.

What You’ll Do

Design and build backend and full-stack systems for AI agents and healthcare automations Build product and platform features for automation creation, workflow pathways, testing, editing, monitoring, and production reporting Improve reliability through better run state management, queues, error classification, alerts, dashboards, and observability Collaborate with product, design, deployment engineering, and customer-facing teams to turn real healthcare workflows into production-grade automation systems Own projects from implementation through deployment, monitoring, iteration, and customer impact Learn from senior engineers while taking on meaningful responsibility early You May Be a Fit If You are graduating soon or recently graduated with a degree in Computer Science, Engineering, Math, or a related technical field You have strong software engineering fundamentals and are excited to keep improving them You have built meaningful projects, internships, research systems, open-source work, startups, or production software You are comfortable with ambiguity and can make progress even when the answer is not obvious You are a self-starter who takes ownership instead of waiting to be handed every next step You move quickly, learn quickly, and care about shipping reliable work You are excited about AI and already use AI tools to improve how you code, debug, learn, or build You want to work in person with a high-caliber team in Toronto Even Better If You have experience with TypeScript, Node, React, Python, or similar backend/full-stack technologies You have worked with databases, APIs, queues, async systems, or distributed workflows You have built side projects, developer tools, automation projects, or AI-powered products You have experimented with LLMs, agents, browser automation, evaluations, or model-powered workflows You have worked in a startup, research lab, internship, hackathon team, or environment where you had to figure things out quickly What You’ll Get Real ownership from day one Mentorship from strong senior technical leaders Training, feedback, and coaching to help you grow quickly Exposure to production AI systems, real customers, and complex healthcare workflows A chance to build technology that directly improves healthcare operations and patient access A high-growth environment where slope matters more than years of experience Why Magical Most new grad jobs optimize for safety, structure, and narrow scope. This is different. At Magical, you will work on hard technical problems, inside a fast-growing company, in one of the most important industries in the world. You will be surrounded by people who expect a lot, move quickly, and care deeply about building something that works. We are still early. The problems are hard. The bar is high. If you are an ambitious new grad who wants to grow fast, take on real ownership, and build AI systems with actual human impact, we would love to meet you.