Invigilo Technologies Pte. Ltd. – C++ Machine Learning Engineer (Computer Vision / Deepstream)

About the Role

We are looking for a highly driven C++ Machine Learning Engineer to help bring our computer vision research into production. You will work closely with our AI research team to convert Jupyter notebook prototypes into optimized, production-grade systems deployed on both edge GPU servers and cloud infrastructure.

This role is ideal for someone who enjoys building real-world ML systems, optimizing inference performance, and working with modern GPU-based deployment stacks such as NVIDIA DeepStream, TensorRT, and other inference frameworks.

This position is supported under the SGInnovate grant, where 75% of the first-year salary will be subsidized. Strong performers will be offered a full-time permanent role in the subsequent year.

Key Responsibilities

• Convert AI research prototypes (Python/Jupyter notebooks) into production-ready C++ applications and services
• Develop and optimize real-time computer vision inference pipelines using NVIDIA DeepStream
• Integrate models into production using TensorRT, ONNX Runtime, Triton Inference Server, or similar frameworks
• Optimize pipelines for performance:
  • FPS throughput
  • GPU utilization
  • latency reduction
  • memory efficiency
• Deploy inference services across both edge devices and cloud environments
• Build robust software components for video processing, multi-camera ingestion, and real-time alerting workflows
• Work with the MLOps and backend engineering teams to integrate pipelines into production systems
• Support debugging and troubleshooting in real deployment environments (edge servers, CCTV systems, GPU machines)
• Maintain clean code standards, documentation, and production-level testing practices

Requirements

• Strong proficiency in C++ with solid software engineering fundamentals
• Experience working with computer vision pipelines (OpenCV, video streaming, RTSP feeds, etc.)
• Prior experience deploying or serving ML models in production environments
• Familiarity with GPU acceleration and inference optimization
• Comfortable working in Linux environments (Ubuntu), Docker, and production deployments
• Strong problem-solving skills and ability to debug performance bottlenecks

Good to Have

• Experience with NVIDIA DeepStream
• Experience with TensorRT, CUDA, GStreamer
• Familiarity with YOLO / object detection models and tracking pipelines
• Experience with Triton Inference Server, Kafka/Redis, REST/gRPC inference services
• Exposure to edge deployment environments and hardware constraints
• Knowledge of Python for bridging research-to-production workflows

What We’re Looking For

• Highly motivated and hardworking individuals
• Comfortable in a fast-paced startup environment
• Objective-focused mindset (results matter more than strict working hours)
• Strong ownership mentality and willingness to take responsibility end-to-end
• Ability to collaborate across research, engineering, and operations teams

Why Join Us

• Work on real-world AI deployments across construction and industrial sites
• Hands-on exposure to large-scale CCTV analytics and edge GPU inference
• Opportunity to build and optimize production systems that directly impact safety outcomes
• High-growth environment with strong career development potential
• Clear conversion pathway to full-time permanent employment after Year 1