Lead Engineer, Applied ML (Remote within Canada)

Company: Confidential — an early-stage, founder-led technology company (full details shared with shortlisted candidates) Location: Canada (EST preferred) · Hybrid, with periodic project-based travel Type: Full-time — also open to 1-year contractors Reports to: Founder / CEO Compensation: Approx. $220–$450k CAD base + equity options for full-time hires About the Role Our client is an early-stage, founder-led company building AI-driven analytics for multi-sensor data, working across commercial, regulatory, and government/defense programs. Their platform has already been validated in real operational environments with early customers, and they're now extending it into new defense, security, and commercial applications. They're hiring a Lead Engineer with applied ML depth. You'll work directly with the founder to build the company's technical capability from the ground up. This is a true foundational role with a lot of autonomy. Your first project extends the platform into a new AI-driven decision-support system for a defence-sector client. It fuses multi-domain sensor data (radio-frequency, infrared, and visual-band), applies ML-based signal classification, and delivers a real-time visualization and decision-support interface. You'll be the primary technical builder across data ingestion, model development, dashboard, and system integration, working alongside a small set of specialist subcontractors (human factors, independent model review, security audit) who validate and stress-test what you build. Past that first project, you'll help shape the company's engineering practices and pick up new work as the project portfolio grows. What You'll Own Data pipeline & fusion Stand up ingestion connectors for a range of sensor and data sources over common transport protocols (e.g. REST, gRPC, MQTT) Define a single unified schema and metadata model; identifiers, timestamps, frequency, location, calibration Align streams across time and space, and add automated quality checks that catch dropouts, outliers, and malformed records Applied ML Build and train both supervised and unsupervised models for signal classification and anomaly detection Tune inference for near-real-time latency. This entails; profiling, quantization, pruning, and similar techniques. Produce the evaluation evidence (accuracy, precision/recall, false-positive rates) needed to support independent third-party model review Visualization & dashboard Build a live, GPU-accelerated dashboard that renders fused data with overlays and per-result confidence scoring. Design adaptive visual layers that stay responsive at near-real-time refresh under full data load. Fold in usability findings from an external human-factors reviewer. Systems integration Bring the pipeline, models, and dashboard together into one modular, containerized system exposed through secure APIs. Support deployment and scenario-based testing inside client test environments. Partner with an external security auditor to close out findings ahead of deployment. Requirements What We're Looking For Strong full-stack engineering background. You are comfortable owning a system end to end. Applied ML experience: building, training, and deploying models in production or near-production settings, not just research or prototyping. Experience with real-time or near-real-time data pipelines and systems integration. Cloud infrastructure experience (AWS preferred, reflecting the current stack) and containerized deployment. Comfortable operating as the sole technical owner, with subcontractor partners handling independent review and validation rather than a peer engineering team. Bonus: signal processing, RF/sensor data, or defense/regulatory technical environments. Bonus: Dual citizenship (Canada & USA).