Embedded Vision ML Engineer for Low-Latency Robotics

Mecka; Toronto; Hybrid; CAD 120,000 - 180,000; Full time

Job summary

Mecka AI in Toronto is hiring to bring advanced computer vision into robotics hardware. You will train models, run them on embedded devices, and ensure latency, memory and power budgets are met.

You will build efficient C++ inference pipelines, optimize runtime performance, and evaluate accuracy across changing conditions. For a lead role, you’ll shape architecture and mentor engineers.

Qualifications

• Experience training and evaluating computer-vision models.
• Experience deploying on embedded Linux and resource-constrained hardware.
• Knowledge of quantization and inference optimization.
• Ability to evaluate accuracy under varying conditions.
• Lead: architecture decisions and guiding engineers.

Responsibilities

• Train and evaluate computer-vision models for embedded devices.
• Build efficient C++ inference pipelines integrated with embedded Linux.
• Optimize models for latency, memory, power.
• Test performance on representative data across hardware.
• Lead: set architecture and evaluation standards.

Skills

Python

C++

PyTorch

Embedded Linux

Quantization

Model evaluation

Performance profiling

Job description

Mecka AI in Toronto is hiring to bring advanced computer vision into robotics hardware. You will train models, run them on embedded devices, and ensure latency, memory and power budgets are met.

You will build efficient C++ inference pipelines, optimize runtime performance, and evaluate accuracy across changing conditions. For a lead role, you’ll shape architecture and mentor engineers.