Computer Vision / AI Engineer

NEDO PAKISTAN, a leading provider of software, hardware, and product development solutions for over 25 years, serves high-profile clients including R&D organizations, OEMs, industrial firms, telecom operators, and government institutions across Pakistan. With its head office in Karachi and branches in Lahore and Islamabad, the company is currently seeking a Mid-Level Computer Vision / AI Engineer to join its TMS Division in Karachi. The role focuses on developing a real-time face enrollment and verification system using existing CCTV/IP cameras, integrating AI capabilities with the current .NET backend and React frontend teams.

Key Responsibilities:
- Develop and integrate face detection and recognition solutions.
- Build face enrollment and verification workflows utilizing CCTV/IP camera streams.
- Work with real-time video streams such as RTSP.
- Implement image and video processing pipelines using Python and OpenCV.
- Adapt and fine-tune pre-trained face recognition models for specific use cases.
- Generate and compare face embeddings for verification purposes.
- Conduct face quality assessments including blur, brightness, face size, and pose.
- Implement or integrate liveness and anti-spoofing solutions.
- Optimize face recognition inference for real-time performance on CPU/GPU platforms.
- Integrate AI services with existing .NET APIs and develop APIs for communication between AI engines and backend systems.
- Test system performance under varying lighting, camera angles, distances, and video quality conditions.
- Continuously measure and improve recognition accuracy and overall system efficiency.
- Maintain comprehensive technical documentation and deployment guidelines.

Required Qualifications:
- 2 to 4 years of relevant experience in Computer Vision, AI, or Machine Learning.
- Strong programming skills in Python.
- Hands-on experience with OpenCV.
- Proficiency in at least one deep learning framework such as PyTorch or TensorFlow.
- Solid understanding of face detection, face recognition, face embeddings, image and video processing, and similarity matching.
- Experience working with pre-trained AI/ML models.
- Basic knowledge of model inference and performance optimization techniques.
- Familiarity with Linux environments.
- Understanding of REST APIs and integration of AI services with backend systems.
- Ability to troubleshoot and optimize real-time video processing pipelines.
- Hands-on project experience in Computer Vision or Face Recognition is essential; candidates with only theoretical knowledge may not be suitable.

Preferred Qualifications and Benefits:
- Experience working with RTSP/IP cameras.
- Familiarity with multimedia frameworks such as FFmpeg or GStreamer.
- Knowledge of ONNX Runtime or TensorRT for model deployment.
- Experience with NVIDIA GPU/CUDA environments.
- Exposure to face recognition libraries and frameworks.
- Practical experience in liveness detection and anti-spoofing techniques.
- Experience with containerization tools like Docker.
- Proven track record of deploying computer vision applications in production environments.

The successful candidate will collaborate closely with existing .NET backend developers, React frontend developers, and the product/project teams. The primary focus will be on the Computer Vision and AI components rather than general frontend or backend development.

Experience:
Candidates should have 2 to 4 years of relevant hands-on experience. Those with around 2 years of strong practical experience in Computer Vision or Face Recognition projects are encouraged to apply.

Job Location:
Karachi, Pakistan (SMCH Shahrah-e-Faisal)

Working Hours:
Monday to Friday: 09:00 AM – 06:00 PM
Alternate Saturdays: 09:00 AM – 05:00 PM

This is an in-person role based at the Karachi office.