Machine Learning/Computer Vision Engineer
Role overview
We are seeking a Machine Learning / Computer Vision Engineer to join our team. You will work on advancing our machine learning capabilities across the full pipeline — from data processing to model development to reporting. This is a hands-on role requiring both research awareness and production-minded engineering.
Responsibilities
- check_circle Design, train, and evaluate classification models for complex visual and geometric data
- check_circle Implement and benchmark modern vision foundation models (DINOv2, CLIP, ViT, ConvNeXt, or similar)
- check_circle Build learned multi-view fusion architectures (e.g., MVCNN) for combining information across multiple perspectives of an object
- check_circle Fine-tune pre-trained vision backbones on domain-specific imagery
- check_circle Develop multimodal models that combine visual features with structured text and attribute data
- check_circle Explore 3D geometry-based classification using point cloud methods (PointNet++, Point Transformer, DGCNN, or similar)
- check_circle Evaluate model performance through rigorous metrics, ablation studies, and iterative experimentation
- check_circle Contribute to data pipeline development, automated reporting, and system productionization
- check_circle Python — strong proficiency
- check_circle PyTorch — model development, custom training loops, fine-tuning, inference
- check_circle Computer Vision — transfer learning, feature extraction, embedding-based methods
- check_circle Vision Foundation Models — hands-on experience with at least one of: DINOv2, CLIP, ViT, ConvNeXt, EfficientNet-V2
- check_circle Multi-view 3D Recognition — familiarity with MVCNN or learned view-pooling techniques
- check_circle ML evaluation — classification metrics, stratified data splitting, experiment design
- check_circle 3D Point Cloud Learning — PointNet, PointNet++, DGCNN, or Point Transformer
- check_circle Multimodal ML — combining vision and text/structured data (cross-attention, fusion architectures)
- check_circle 3D data formats & tools — STEP, IGES, B-Rep; Open3D, trimesh, FreeCAD, Creo Parametric or SolidWorks
- check_circle CAD-native learning — awareness of UV-Net, BRepNet, or DeepCAD
- check_circle MLOps — experiment tracking (MLflow, W&B), model versioning, CI/CD
- check_circle Manufacturing/engineering domain knowledge — part taxonomies, attribute systems
- check_circle Experience productionizing research-stage ML code (packaging, testing, configuration, logging)
- check_circle GPU/CUDA environment setup and management
- check_circle BS/MS in Computer Science, Machine Learning, Computer Vision, Data Science, Engineering, or equivalent practical experience
About the company
ITC Infotech is a leading global technology services and solutions provider, led by Business and Technology Consulting. ITC Infotech provides business-friendly solutions to help clients succeed and be future-ready, by seamlessly bringing together digital expertise, strong industry specific alliances and the unique ability to leverage deep domain expertise from ITC Group businesses. We provide technology solutions and services to enterprises across industries such as Banking & Financial Services, Retail, Healthcare, Manufacturing, Consumer Goods, Travel and Hospitality, through a combination of traditional and newer business models, as a long-term sustainable partner.