Software Engineer, Data Science

Toronto 4 days agoFull-time External
Negotiable
Role Overview Trajekt is seeking a Software Engineer to join our growing data science team. This role involves significant ownership of the core technology that controls our robots, support for a wide variety of internal R&D projects across the company, and the design of tools to facilitate data-informed decision-making for both engineering teams and executive business functions. Your primary role will be to build maintainable, production-grade systems that deliver our data science tools, robot control models, and physics prediction models. You will work in a team with deep expertise in mathematics and physics. Your role is to complement that strength with strong software design, code quality, tooling, and operational rigor, while still being comfortable engaging directly with these topics. The data science team at Trajekt is core to supporting other teams with complex mathematical problems. If you have strong mathematical and data science skills, but consider yourself a software engineer first and foremost, then you are likely a great fit for this role. Trajekt is a uniquely strong place to build a career because it combines: A high degree of ownership early in your tenure A profitable, well-capitalized company with stable customers and revenue Competitive compensation Opportunities for career growth, both vertically and horizontally A tight-knit team of domain experts who truly enjoy working together You'll have the space to do deep work, as well as the support to grow quickly in the way that best suits you and the stability to think long-term. Key Responsibilities Software Engineering for Data and Models Own and improve the Python codebases used for training, validating, and deploying models Design clean APIs and abstractions around mathematical models and physical simulations Refactor research outputs into maintainable, testable, production-ready systems Establish patterns for configuration, logging, testing, deployment, and reproducibility Model Deployment & Maintaining Production Systems Support the full lifecycle of models, from research to deployment Build and maintain tooling for running models in real time and analyzing outcomes Ensure numerical stability, performance, and correctness in production environments Cross-Company Support Partner with teams across robotics, computer vision, web, and business operations Provide software and analytical support for complex mathematical or data-driven problems Build internal tools that make advanced analysis accessible to non-specialists Support exploratory R&D projects with interdisciplinary teams R&D and Physical System Modelling Assist in the design of novel methods to model robot dynamics, ball flight aerodynamics, and other physical systems Collaborate on experimental design, data collection, and analysis Required Experience & Skills 3+ years of professional software development experience Strong Python proficiency, including: Modular code design and package structure Testing, debugging, and refactoring large codebases Performance-aware numerical programming Experience in maintaining production systems used by other engineers Comfort working with mathematically complex code Experience collaborating with researchers, data scientists, and other domain experts Strong engineering and software development judgment and intuition Nice to Have Experience with scientific computing libraries (NumPy, SciPy, PyTorch, JAX, etc.) Exposure to physics-based modeling or simulation Familiarity with data pipelines, experiment tracking, or model evaluation workflows Experience working with robotics, hardware-adjacent software, or real-time systems Experience with computer vision Experience leading teams of engineers Comfort reading academic-style code or research papers and translating them into software We are looking for someone who is eager to take ownership, work closely with others, and get things done while having fun through the process. The ideal candidate has a track record of taking pride in building systems that others can rely on.