Computer Vision Engineer -DSA

About vConstruct

Role Overview

Computer Vision Development

• Design and develop computer vision solutions for construction and engineering applications.
• Build and optimize models for image and video analytics, visual inspection, object detection, segmentation, OCR, tracking, and quality assessment.
• Collaborate with AI engineers and domain experts to deliver scalable production-ready vision solutions.

AI Model Development

• Build and train CNN, Vision Transformer, DETR, YOLO, Mask R-CNN, Segment Anything, Grounding DINO, CLIP, Florence, and related architectures.
• Fine-tune foundation vision models and apply transfer learning and self-supervised learning.
• Build multimodal AI systems combining vision and language models.

Dataset Engineering

• Design and manage image and video datasets for computer vision applications.
• Develop annotation strategies and ensure high-quality labeled datasets.
• Apply data augmentation, synthetic data generation, and active learning to improve model performance.

Research and Innovation

• Evaluate current computer vision research and experiment with emerging vision foundation models.
• Prototype innovative AI solutions and publish technical findings internally.

Required Technical Skills

• Strong Python programming skills with AI and machine learning development experience.
• Hands-on experience with deep learning and computer vision frameworks.
• Experience with object detection, image classification, segmentation, OCR, video analytics, and visual inspection.
• Experience with CNNs, Vision Transformers, foundation vision models, image and video processing, model training, evaluation, and optimization.
• Experience preparing datasets through annotation, augmentation, and quality validation.
• Familiarity with GPU-based training and inference optimization.
• Experience integrating AI models into production applications and collaborating with software engineering teams.

Preferred Experience

• Working knowledge of vision-language models and multimodal AI is preferred.
• Four to eight years of computer vision experience is preferred.
• Production AI deployment, custom deep learning model training, GPU optimization, large-scale image or video dataset experience, and Agile product team experience are preferred.