Agrex.ai is a video analytics company transforming existing CCTV infrastructure into intelligent, real-time monitoring systems for retail, manufacturing, banking, logistics, and education. Our platform delivers people & vehicle analytics, SOP/compliance monitoring, and operational insights--on edge devices (e.g., Jetson) and in the cloud.
Role Overview
We're looking for a Computer Vision Intern who's excited to learn and contribute to real-world video analytics systems. You'll get hands-on experience with training and optimizing models, building video pipelines, and deploying on Jetson and NVIDIA GPUs--while working closely with our Product and Engineering teams.
What You'll Do
Assist in implementing, training, and evaluating detection/segmentation/tracking models (YOLO, RTDETR, Mask models, ByteTrack, OC-SORT).
Help integrate models into NVIDIA DeepStream / GStreamer pipelines; learn to export to ONNX/TensorRT and optimize for speed & efficiency.
Contribute to features like people/vehicle counting, dwell/queue analytics, intrusion detection, PPE checks, and simple re-ID.
Work with real video streams (RTSP/H.264/H.265)--frame sampling, jitter handling, and reconnection logic.
Write clean Python code (and explore some C++ for performance); assist with unit tests and benchmarking.
Support dataset preparation: curation, labeling guidelines, using CVAT/Label Studio, and basic augmentation.
Track experiments with metrics (mAP, recall, FPS, GPU utilization) and document learnings.
Debug and test field issues with the team; contribute to prototypes and quick fixes.
Learn best practices for handling enterprise video data securely.
What We're Looking For
Good coding skills in Python (NumPy, OpenCV, PyTorch/TensorFlow).
Understanding of basic Computer Vision/ML concepts: convolutions, IoU, NMS, augmentation, overfitting, metrics.
Exposure to at least one area: object detection, segmentation, or tracking.
Comfortable learning from documentation, research papers, and implementing small utilities/tools.
Bonus Points (Nice to Have)
Familiarity with NVIDIA stack: CUDA, TensorRT, DeepStream.
Hands-on with Jetson (Xavier/Orin) or other edge devices.
Knowledge of GStreamer/FFmpeg and video codecs.
Experience with DVC, W&B, Docker, or ML pipelines.
Interest in re-ID, pose estimation, OCR, or multi-camera systems.
Previous projects/internships related to retail, factory, warehouse, or surveillance data.
Education
B.Tech/BE/M.Tech in CS/EE/Maths or equivalent practical experience
Job Type: Full-time
Pay: ?15,000.00 per month
Work Location: In person
Beware of fraud agents! do not pay money to get a job
MNCJobsIndia.com will not be responsible for any payment made to a third-party. All Terms of Use are applicable.