Edge ML Engineer (Full Stack)

Voxelis is an airborne AI company on a mission to save lives and protect the environment. We build intelligent systems that fly on real aircraft, processing sensor data in real time to support wildfire response, search and rescue, environmental monitoring, and other critical dual-use applications. Our technology operates at the edge, on full-scale, certified aircraft, in real-world missions. We work across government and commercial sectors, and we’re looking for people who are energized by building technology that has tangible, real-world impact. If the idea of your code running on a helicopter over an active wildfire excites you, keep reading.

ABOUT THE ROLE

We’re hiring an Edge ML Engineer to own the multi-model data pipeline and catalog infrastructure for our VoxVision platform. This is a hands-on role spanning the full stack: from training and optimizing neural networks, to deploying them on embedded hardware in the air, to building the cloud-side data infrastructure that stores, catalogs, and makes all of that data searchable. You'll also build multi-model pipelines where detection, classification, and tracking models operate autonomously in sequence and in parallel. This isn’t a desk-only job; you’ll have opportunities to fly with the systems you build, see your models run in the air, and work directly with operational teams in the field. If you’re the kind of engineer who wants to build models from scratch and also cares deeply about how data flows through a system end to end, this role is for you.

WHAT YOU’LL DO

Pipeline Architecture & Implementation

• Architect an end-to-end data pipeline that handles raw sensor ingestion, model inference, and derived product generation across airborne, ground, and cloud tiers.
• Build pipeline components on the VoxVision platform that autonomously orchestrate multiple AI models in sequence and in parallel, producing outputs such as detections, classifications, segments, and tracks.
• Implement edge processing logic for intelligent data filtering, compression, and selective retention to manage bandwidth constraints before cloud transmission.
• Ensure pipeline robustness under real-world operational conditions, including intermittent connectivity and low-bandwidth scenarios.

Edge ML & Model Development

• Build and train ML models from scratch using PyTorch and TensorFlow, with a focus on detection, classification, and segmentation tasks relevant to airborne sensing and remote observation.
• Optimize and deploy models to edge hardware using TensorRT and equivalent acceleration frameworks for real-time inference.
• Own the data preparation pipeline: cleaning, annotation, curation, and quality assurance of training and evaluation datasets on backend infrastructure.

Data Catalog & Metadata Infrastructure

• Design and implement the data infrastructure to store raw sensor data and derived products alongside a rich metadata catalog, including schema design, indexing strategies, and retention policies.
• Build automated services that use model outputs to tag and index data by time, location, platform, object type, confidence level, and other mission-relevant attributes.
• Develop search and query capabilities (APIs and tooling) that allow users to retrieve events, objects, or scenes by time, geography, platform, and object-level attributes.
• Validate catalog searchability and query performance under representative operational loads.

REQUIRED QUALIFICATIONS

Education

• Bachelor’s degree in Computer Science, Software Engineering, Mathematics, Electrical Engineering, or a closely related quantitative discipline (or equivalent demonstrated experience). Master’s/Ph.D. a plus.

Technical Experience

• Experience with computer vision tasks including object detection, classification, segmentation, and tracking applied to real-world sensor data
• ML frameworks and libraries, with demonstrated ability to build custom model architectures from scratch.
• Proficiency in Python and at least one systems-level language (C++, Rust, or Go).
• Familiarity with model optimization (TensorRT) and deployment to edge hardware (e.g., NVIDIA Jetson, Intel Movidius, or similar embedded platforms).
• Experience developing edge-to-cloud data pipelines with real-time or near-real-time processing requirements.
• Experience with data cleaning, annotation workflows, and dataset curation at scale.

PREFERRED QUALIFICATIONS

• Experience with AWS services (e.g., S3, Lambda, SQS, ECS, IoT Core) or equivalent cloud platforms for backend data infrastructure.
• Experience with containerization (Docker, Kubernetes) and CI/CD for ML workloads.
• Familiarity with database technologies (SQL, NoSQL, time-series) and metadata/catalog management patterns.
• Experience with geospatial data standards (GeoJSON, STAC, OGC services) and spatiotemporal indexing.
• Familiarity with ROS2 and message brokers (MQTT, RabbitMQ, Kafka)
• Comfort working up and down the stack, from low-level model optimization on embedded hardware to backend API development and cloud infrastructure.
• Familiarity with airborne sensor systems, remote sensing data, or related sensor modalities (EO/IR, radar, lidar).
• Experience operating systems under constrained network conditions (intermittent, low-bandwidth, high-latency).

WHAT WE OFFER

• Work on AI systems that directly support wildfire response, disaster relief, environmental protection, and other high-impact dual-use applications. Your work will have visible, meaningful impact.
• Opportunities to fly with your systems and see your models operate in real-world airborne missions.
• Join a team at the forefront of deploying AI to airborne platforms. We’re not researching this in a lab, we’re flying it.
• Our compensation packages include competitive base salary and equity in the form of stock options.
• Small, high-performing team where your work ships fast and your contributions are felt immediately.
• Access to airborne inference compute and real-world mission datasets - not just benchmark data.
• Professional development and conference attendance support.
• Hybrid-flexible with a preference for regular on-site collaboration.

EQUAL OPPORTUNITY

Voxelis is an equal opportunity employer. We celebrate diversity and are committed to creating an inclusive environment for all employees. All qualified applicants will receive consideration for employment without regard to race, color, religion, sex, sexual orientation, gender identity, national origin, disability, or veteran status.

Job Type: Full-time

Pay: $95,000.00-$130,000.00 per year

Application question(s):

• Are you able to work on-site/hybrid in Richmond, BC or Calgary, AB?
• Have you built and deployed an ML model to production? If yes, describe in one sentence what it did and where it ran.
• This role involves deploying models to edge/embedded hardware. How familiar are you with this?
• What interests you about this role specifically?

Work Location: Hybrid remote in Richmond, BC V7B 1K3