Top Tech Jobs & Startup Jobs in Montreal

Architect, build, deploy, and operate production-grade AI agentic systems and workflows. Responsibilities include backend and full-stack development, agent orchestration, workflow state, retrieval, memory, approvals, failure recovery, data architecture, enterprise integrations, cloud infrastructure, observability, security, and scaling. The role establishes reusable engineering patterns and partners with users to deliver reliable AI products, while addressing production issues involving model behavior, permissions, latency, cost, and consistency.

Develop and research deep learning and vision algorithms for autonomous driving (detection, tracking, prediction, planning, SLAM). Lead end-to-end ML projects from data analysis through model evaluation, and collaborate across product, simulation, and algorithm teams to extend ML technologies into system components.

Build high-performance, scalable simulation infrastructure and tooling for autonomous vehicle development. Design systems for scenario generation, simulation execution, synthetic data, and reinforcement-learning workflows. Collaborate with ML, perception, and controls teams to improve realism, performance, and validation pipelines while writing production-quality C++ and Python code, contributing to architecture, testing, and documentation.

Develop and train conditioned policies and MARL systems to simulate realistic driving behaviors, implement safety-constrained RL algorithms, design rewards and evaluation metrics, optimize large-scale training pipelines, advance neural architectures for long-horizon planning and spatial reasoning, and integrate research models with production simulation and planning teams.

The Senior ML/RL Engineer will develop behavior models, implement RL algorithms focusing on safety, design reward functions and optimize training environments while collaborating across teams.

Optimize end-to-end GPU performance for real-time autonomous driving workloads: develop and optimize parallel GPU algorithms and inference pipelines, profile and eliminate bottlenecks across computation and memory, debug GPU software for latency/throughput improvements, and collaborate on onboard GPU software architectures for embedded platforms.