AI Engineer – Computer Vision | PyTorch | TensorRT

Experience: 3–6 Years
Salary: Up to ₹12 LPA
Work Mode: On-site
Education: BE/BTech/MTech/MS in Computer Science, AI/ML, ECE or related fields

Job Description

We are looking for an experienced AI Engineer with strong hands-on expertise in computer vision, deep learning, real-time object detection and GPU inference optimization. The role involves developing and deploying detection and tracking models for EO/IR (Electro-Optical/Infrared) surveillance and real-time video analytics systems.

Key Responsibilities

• Train and fine-tune object detection models using PyTorch and custom datasets.
• Deploy optimized models using ONNX and NVIDIA TensorRT.
• Optimize real-time inference performance, FPS, latency and GPU utilization.
• Implement multi-object tracking using ByteTrack or similar algorithms.
• Work on detection accuracy, false alarms, missed detections and tracking failures.
• Process multiple video streams and improve inference efficiency.
• Evaluate model performance using mAP, precision, recall and operational metrics.
• Debug model deployment, preprocessing and GPU runtime issues.
• Maintain model versions, evaluation reports and regression test results.

Mandatory Skills

• 3+ years of hands-on experience in PyTorch-based object detection and computer vision.
• Experience training or fine-tuning YOLO or other modern detection models.
• Practical experience deploying models using ONNX and TensorRT, or an equivalent optimized inference runtime.
• Strong knowledge of CUDA/GPU inference and performance optimization.
• Experience with multi-object tracking algorithms such as ByteTrack or DeepSORT.
• Strong Python programming and production deployment experience.
• Understanding of FPS, inference latency, precision, recall and model evaluation.

Preferred Skills

• Experience with EO/IR, infrared or thermal image datasets.
• NVIDIA Jetson/JetPack and edge AI deployment.
• Real-time video analytics or surveillance systems.
• GPU profiling using NVIDIA Nsight.
• Defence, robotics, autonomous systems or industrial vision applications.

Important

Candidates must have practical model development and deployment experience. Candidates with only academic projects, basic AI/ML knowledge, or GenAI/LLM experience without computer vision and object detection are not suitable.

Interested candidates can apply with their updated CV highlighting relevant computer vision projects, TensorRT/ONNX deployment experience, GPU performance metrics, current CTC, expected CTC and notice period.

Pay: ₹350,000.00 - ₹1,100,000.00 per year

Work Location: In person