Autonomous Robotics: Multimodal AI Engineer (Hybrid)

Asiacruit BPO, Inc.; United States; Hybride; USD 120 000 - 180 000; Plein temps

Résumé du poste

Asiacruit invites applications for a Multimodal AI Engineer to architect Vision-Language-Action models for autonomous robotic systems. You will refine perception, planning, and action-generation modules across edge deployments and collaborate with AI researchers and robotics engineers to translate research into practical AI capabilities.

The role focuses on fine-tuning multimodal models, building multi-camera perception pipelines, and optimizing models for deployment with state-of-the-art

Qualifications

• Master's degree or Ph.D. in Computer Science, AI, Robotics, ML, CV, or related field.
• Approximately 3–5 years of relevant experience in multimodal AI, deep learning, CV, foundation models, or robotic learning.
• Strong programming skills in Python and hands-on experience with PyTorch.
• Experience with Transformer-based architectures and multimodal models.
• Understanding of vision-language models and multimodal representation learning.
• Familiarity with LoRA or other PEFT techniques.

Responsabilités

• Develop, fine-tune, and adapt ViTs and multimodal models for robotic perception and action generation.
• Implement action-chunking architectures to support long-horizon planning.
• Build multi-camera fusion pipelines for VLA perception and decision-making.
• Develop multimodal feature extraction pipelines linking perception to planning and action.
• Integrate computer vision capabilities (face recognition, perimeter threat assessment) into multimodal workflows.
• Apply PEFT techniques like LoRA to adapt models.
• Work with transformers, visual embeddings, and vector search to improve multimodal representations.
• Profile and optimize models via pruning and other compression methods.
• Prepare model weights for deployment using NVIDIA TensorRT and related tools.
• Evaluate model performance through experiments and benchmarking.
• Collaborate with researchers across VLA, robotics controls, and edge AI.

Connaissances

Python

PyTorch

Transformer architectures

Vision Transformers

PEFT (LoRA)

Edge AI

Vector search

Formation

Master's degree in CS/AI/Robotics

Ph.D. in a related field (preferred)

3–5 years multimodal AI/robotics experience

Outils

NVIDIA TensorRT

NVIDIA DeepStream

PyTorch

Description du poste

Asiacruit invites applications for a Multimodal AI Engineer to architect Vision-Language-Action models for autonomous robotic systems. You will refine perception, planning, and action-generation modules across edge deployments and collaborate with AI researchers and robotics engineers to translate research into practical AI capabilities.

The role focuses on fine-tuning multimodal models, building multi-camera perception pipelines, and optimizing models for deployment with state-of-the-art