Computer Vision EngineerBlock 9 Consultancy • Vancouver, British Columbia, Canada
Dieses Stellenangebot ist nicht mehr verfügbar
Computer Vision Engineer
Block 9 Consultancy
- Vancouver, British Columbia, Canada
- Vancouver, British Columbia, Canada
Über
We're hiring a Computer Vision Engineer to build and deploy vision models that run reliably on edge hardware (NVIDIA Jetson) in real-world conditions. This is a delivery-focused, high-ownership role: you will take models from data → training → optimization → edge deployment → monitoring, with clear targets for accuracy, speed, stability — and personal responsibility for getting systems into production.
This is not a research-only role. You will be accountable for systems that run in the field, under messy, real-world conditions, and you'll be expected to drive work forward without being micromanaged.
What You'll Work On
Real-world computer vision tasks such as object detection, counting, tracking, and brand/item recognition using deployed camera feeds
Edge inference on NVIDIA Jetson devices, including performance tuning and deployment reliability
Robust performance in uncontrolled environments: lighting variation, occlusions, motion blur, dirty/partial views, and camera angle changes
Key Responsibilities
Design, train, and improve computer vision models (detection, classification, segmentation, tracking as needed)
Build rigorous evaluation workflows: accuracy metrics, failure-mode analysis, and repeatable test sets
Own the edge deployment pipeline: package models, manage versions, and ensure consistent inference behavior on Jetson devices
Optimize inference performance (e.g., ONNX/TensorRT, quantization where appropriate) to meet latency and throughput targets
Collaborate closely with software and hardware teams to integrate inference into the full system (data flow, APIs, logging, device constraints)
Diagnose production issues using logs, sample captures, and structured root-cause analysis
Create clean, reliable documentation: model versions, evaluation results, deployment steps, and rollback plans
Take ownership of production outcomes — when systems fail, you drive the fix to resolution
Required Qualifications
Strong Python skills and solid software engineering fundamentals (Git, testing, clean code practices)
Deep learning experience using PyTorch or TensorFlow
Practical computer vision experience (image/video pipelines, data augmentation, OpenCV or equivalent)
Proven ability to ship beyond notebooks: deploying, integrating, and maintaining models in production systems
Strong communication skills and ability to clearly explain technical trade-offs
Strong Pluses
Hands-on NVIDIA Jetson experience (Linux, CUDA basics, Docker, device-level debugging)
Model optimization experience (ONNX, TensorRT, quantization, profiling, GPU utilization tuning)
Real-time video pipeline experience (multi-camera feeds, frame drops, buffering, synchronization)
Experience designing data workflows: labeling guidance, dataset QA, dataset/version control
Mindset & Operating Style (This Role Is Not for Everyone)
This is a high-ownership, delivery-driven role in a fast-moving startup environment. We are looking for builders, not passengers.
Strong ownership mindset — you take problems to completion, not just report blockers
Delivery-first mentality — you prioritize shipping reliable systems over endless experimentation
High accountability — you manage your own deadlines and follow through without being chased
Bias toward action — you propose solutions and initiate fixes, not just identify issues
Comfortable operating with urgency when deadlines demand it
Takes the role seriously and treats production systems with professional care
Growth-oriented — motivated to grow with the company, not just hold a title
If you prefer narrow scope, slow timelines, or low operational responsibility, this role is unlikely to be a fit.
What We Will Screen For
Evidence you can ship: prior deployments, architecture walkthroughs, portfolio/GitHub, or measurable outcomes
Ability to handle messy data and real-world failure cases (not just benchmark results)
Clear communication on trade-offs: accuracy vs latency vs compute cost vs reliability
Ownership behavior in past roles: examples of taking responsibility for delivery, fixes, and production stability
Deadline discipline and ability to deliver under ambiguity
Equal Opportunity
We are an equal opportunity employer and evaluate candidates based on mindset, attitude, skills, experience, and role requirements.
If you're excited by ownership, shipping real systems, and being accountable for production outcomes, we'd love to hear from you. If you prefer narrow scope, slow timelines, or low operational responsibility, this role is unlikely to be a fit.
Sprachkenntnisse
- English
Hinweis für Nutzer
Dieses Stellenangebot wurde von einem unserer Partner veröffentlicht. Sie können das Originalangebot einsehen hier.