Ultrasonic imaging platform for industrial asset inspection and integrity analysis
DarkVision builds acoustic-based imaging hardware and software for detecting defects in critical industrial infrastructure. The tech stack reveals a production-grade ML ops backbone—PyTorch, JAX, SageMaker, Kubeflow, Prefect, DVC—paired with low-level embedded systems work (CUDA, C++, Embedded Linux), suggesting they're shipping both real-time sensor processing and cloud-based analytics. Active hiring skews heavily toward engineering and data (21 of 26 roles), while pain points cluster around manual workflows and dataset handling, indicating they're automating inspection pipelines at scale.
DarkVision designs and manufactures ultrasonic imaging systems for inspecting tubular and pressure vessel infrastructure, serving industries where asset failure carries high safety or financial cost. The company owns the full stack: sensor arrays, high-speed electronics, imaging algorithms, and cloud analytics. Founded in 2013 and based in North Vancouver, DarkVision operates across Canada and the United States with approximately 201–500 employees. Their platform captures ultrasound datasets at industrial scale and converts them into actionable integrity assessments through ML-driven visualization and reporting tools.
DarkVision develops acoustic-based imaging systems for inspecting critical industrial assets like pipelines and pressure vessels. The platform captures ultrasound data, processes it with ML analytics, and generates visual reports on structural integrity.
Core languages: Python, C++, CUDA. ML frameworks: PyTorch, JAX. Infrastructure: AWS (SageMaker, Lambda, Batch), Kubernetes, Docker. Orchestration: Prefect, Apache Airflow, Kubeflow. Design tools: SolidWorks, Fusion 360, Inventor, Creo. ML ops: Weights & Biases, DVC, MLflow.
Other companies in the same industry, closest in size