Multimodal AI Engineer Robotics And Foundation Models

AsiaCruit, Inc.; New Delhi; On-site; INR 3,500,000 - 5,500,000; Full time

Job summary

AsiaCruit is seeking a Multimodal AI Engineer to develop and adapt Vision-Language-Action foundation models for autonomous robotic systems. You will fine-tune ViTs, build multi-camera perception pipelines, and implement action-generation architectures for edge AI deployment.

You will collaborate with AI researchers and robotics engineers to translate research concepts into practical high-performance capabilities, including optimization, benchmarking, and integration with downstream planning

Qualifications

• Master's or PhD in CS/AI/Robotics/ML/CV or related field.
• 3–5 years of relevant multimodal AI, deep learning, computer vision, or robotic learning experience.
• Strong programming in Python with hands-on PyTorch.
• Experience with Transformer-based architectures and Vision Transformers.
• Knowledge of vision-language models and multimodal data processing.
• Experience with PEFT methods such as LoRA.
• Familiarity with robotic learning, action generation, ACT, and diffusion policies.
• Experience with multi-camera perception and feature fusion.
• Understanding of model profiling, pruning, quantization, and optimization.
• Experience with NVIDIA TensorRT and GPU-accelerated inference is a plus.
• Knowledge of embeddings and vector search is a plus.

Responsibilities

• Develop, fine-tune, and adapt ViTs and multimodal foundation models for robotic perception and action generation.
• Implement action-chunking architectures (ACT, Diffusion Policy) for long-horizon planning.
• Build multi-camera fusion pipelines for VLA perception and decision-making.
• Develop and optimize multimodal feature extraction pipelines for robotic planning.
• Integrate CV capabilities like face recognition into multimodal workflows.
• Apply PEFT techniques (LoRA) to adapt models for robotic applications.
• Work with Transformer architectures, visual embeddings, and vector search for multimodal representations.
• Profile and prune models to improve inference efficiency.
• Prepare model weights for deployment using TensorRT and related tools.
• Evaluate performance through experiments, benchmarking, and iterative improvements.
• Collaborate with researchers and engineers across VLA, robotics controls, and edge AI.

Skills

Python

PyTorch

Vision Transformers

Multimodal models

LoRA / PEFT

Reinforcement learning

Edge AI

Education

Master's or PhD in CS/AI/Robotics/ML/CV

Tools

TensorRT

NVIDIA DeepStream

Vector search

Job description

Job Description:

About Asiacruit

At Asiacruit, we connect top talent with forward-thinking organizations across industries. Our mission is to help businesses grow through smart, strategic, and people-focused solutions. We support companies by providing high-quality talent for local and global markets.

Position Overview

We are looking for a Multimodal AI Engineer to develop and adapt Vision-Language-Action (VLA) foundation models for autonomous robotic systems. You will work at the intersection of multimodal perception, foundation models, and robotic learning, helping AI systems interpret visual inputs, plan complex tasks, and translate instructions into physical actions.

This role focuses on fine-tuning multimodal models, developing multi-camera perception pipelines, implementing action-generation architectures, and optimizing models for deployment on edge AI platforms. You will collaborate with AI researchers and robotics engineers to turn research concepts into practical, high-performance AI capabilities.

Key Responsibilities

• Develop, fine-tune, and adapt Vision Transformers (ViTs) and multimodal foundation models for robotic perception and action generation.

• Implement action-chunking architectures, including Action Chunking with Transformers (ACT) and Diffusion Policy, to support long-horizon robotic task planning.

• Build multi-camera fusion pipelines that combine visual inputs into coherent representations for VLA perception and decision-making.

• Develop and optimize multimodal feature extraction pipelines that integrate visual perception with downstream robotic planning and action-generation systems.

• Integrate computer vision capabilities, including face recognition and perimeter threat assessment, into multimodal AI workflows.

• Apply parameter-efficient fine-tuning (PEFT) techniques, including LoRA, to adapt foundation models for specific robotic applications.

• Work with Transformer architectures, visual embeddings, and vector search techniques to improve multimodal representation and information retrieval.

• Profile and optimize multimodal models through pruning and other model compression techniques to improve inference efficiency.

• Prepare and optimize model weights for downstream deployment using NVIDIA TensorRT and related inference optimization tools.

• Evaluate model performance through systematic experiments, benchmarking, and iterative improvements.

• Collaborate with researchers and engineering teams across VLA, robotics controls, and edge AI to support end-to-end system development.

Qualifications

• Masters degree or Ph.D. in Computer Science, Artificial Intelligence, Robotics, Machine Learning, Computer Vision, or a related field.

• Approximately 3–5 years of relevant experience in multimodal AI, deep learning, computer vision, foundation models, or robotic learning.

• Strong programming skills in Python and hands-on experience with PyTorch.

• Experience developing, training, or fine-tuning Transformer-based architectures, Vision Transformers, or multimodal deep learning models.

• Understanding of vision-language models, multimodal representation learning, and the integration of visual information into AI decision-making systems.

• Practical experience with parameter-efficient fine-tuning methods, such as LoRA or other PEFT techniques.

• Familiarity with robotic learning and action-generation approaches, including ACT, Diffusion Policy, imitation learning, or reinforcement learning.

• Experience with multi-camera perception, visual feature fusion, or related computer vision pipelines.

• Understanding of model profiling, pruning, quantization, or other model optimization techniques.

• Familiarity with NVIDIA DeepStream, TensorRT, or GPU-accelerated inference is an advantage.

• Knowledge of embeddings, vector search, and multimodal data processing is a plus.

• Strong analytical and problem-solving skills, with the ability to conduct experiments and translate research findings into practical engineering solutions.

Why Join?

• Work on emerging technologies at the intersection of multimodal AI, foundation models, and robotics.

• Develop AI systems that connect visual perception, intelligent planning, and physical execution.

• Explore advanced approaches to VLA model adaptation, action generation, and multimodal learning.

• Contribute to real-world AI applications involving autonomous systems and intelligent perception.

• Collaborate with researchers and engineers working on challenging AI research and development projects.