Edge AI / MLOps Engineer

Easycomm Innovations; United States; Remote; USD 120,000 - 170,000; Full time

Job summary

Easycomm Innovations is hiring a software engineer to design a computer-vision system that monitors production and dispatch at a large edible oil, soap, and soya meal plant. You will develop edge AI models that process video feeds from industrial cameras installed at exit gates and loading docks.

You’ll build inference pipelines to count items, verify shipments against ERP orders, and trigger safety notifications when discrepancies occur.

Qualifications

• Experience with computer vision frameworks (TensorFlow, PyTorch, OpenCV).
• Hands-on MLOps experience: versioning, testing, deployment pipelines, monitoring.
• Experience optimizing models for edge devices / resource-constrained hardware.
• Ability to work with real-world, messy data in factory environments.

Responsibilities

• Develop edge AI models processing video feeds from industrial cameras at gates and loading docks.
• Build inference pipelines to count items and match shipments to ERP orders.
• Deploy and optimize models on edge servers with low latency.
• Design data pipelines from cameras to central verification service and dashboards.

Skills

TensorFlow

PyTorch

OpenCV

MLOps

Edge optimization

Messy data handling

Job description

Batch: 2024/2025/2026/2027.

About the role

• Engineer a computer-vision system that monitors production and dispatch operations for a large edible oil, soap, and soya meal manufacturer
• Develop edge AI models that process video feeds from industrial cameras installed at exit gates and loading docks
• Build inference pipelines that count items produced, match shipments against ERP orders, and trigger safety notifications when discrepancies occur
• Deploy and optimize machine learning models to run on edge servers with minimal latency and resource constraints

What you'll work on

• Training and fine-tuning vision models to handle challenging factory conditions such as worn or damaged packaging, similar product variants, wrapped pallets, high-speed conveyor lines, dim lighting, and unstable network connectivity
• Designing the data pipeline from industrial cameras through edge processing to a central verification service
• Building APIs and integrations between the edge AI layer, a web dashboard for operators, and mobile applications for dispatch teams
• Monitoring model performance in production and retraining as needed when accuracy drifts due to environmental changes or new SKUs

What we're looking for

• Strong experience with computer vision frameworks such as TensorFlow, PyTorch, or OpenCV
• Proven background in MLOps: model versioning, testing, deployment pipelines, and monitoring
• Hands-on knowledge of edge computing platforms and optimizing models for resource-constrained hardware
• Ability to work with real-world, messy data and adapt solutions when field conditions differ from lab assumptions