Asiacruit
Asiacruit
51 – 200 Employees
B2BRecruitment
Asiacruit is a B2B recruitment and workforce services company that helps businesses hire and manage employees across Asia. Its Employer of Record and Business Process Outsourcing services cover payroll administration, HR support, compliance, and regional expansion without requiring clients to establish local legal entities. Asiacruit also advertises job opportunities and supports cross-border workforce management, with a company culture recognized as a Great Place to Work in the Philippines.

Lead AI Research Scientist, Robotics

Lead research on vision-language-action models for robotic navigation and manipulation. Guide multimodal foundation models from adaptation through deployment.

Description

  • Lead the adaptation and development of vision-language-action models for robotics.
  • Fine-tune open-weight models such as OpenVLA, Octo, and π0 for navigation and manipulation.
  • Build multimodal representations that connect camera input, natural-language instructions, spatial reasoning, and robot actions.
  • Develop action tokenization and spatial waypoint generation methods for robot behavior.
  • Create parameter-efficient fine-tuning pipelines using LoRA and other PEFT methods.
  • Apply vision-language models, Vision Transformers, and large multimodal models.
  • Integrate visual perception and AI into robotic decision-making systems.
  • Investigate cross-embodiment learning and datasets such as Open X-Embodiment.
  • Work with robotics controls and embedded systems teams to turn model outputs into executable robot behaviors.
  • Lead research planning, technical evaluations, experiments, and model performance improvements.
  • Mentor researchers and engineers on model adaptation, training, evaluation, and deployment.

Requirements

  • Hold a master's or doctoral degree in computer science, artificial intelligence, robotics, machine learning, or a related field.
  • Bring at least four years of relevant experience in multimodal AI, foundation models, vision-language models, imitation learning, or robotic learning.
  • Have strong experience developing, training, or fine-tuning large multimodal models, Vision Transformers, or related deep learning architectures.
  • Be proficient in Python and PyTorch or JAX.
  • Have hands-on experience with vision-language models or vision-language-action architectures.
  • Know open-weight VLA models such as OpenVLA, Octo, or π0.
  • Have experience with Hugging Face tools and model training or fine-tuning workflows.
  • Know parameter-efficient fine-tuning techniques, including LoRA or similar methods.
  • Understand multimodal representation learning, visual perception, and language-conditioned action generation.
  • Be familiar with robotic imitation learning, policy learning, or reinforcement learning.
  • Cross-embodiment robotic learning and datasets such as Open X-Embodiment are an advantage.
  • Be able to lead technical research, assess experimental results, and turn research findings into practical AI capabilities.

Benefits

  • Lead development of emerging Physical AI and vision-language-action technologies.
  • Work across multimodal AI, foundation models, and robotics.
  • Explore new methods for language-conditioned robot navigation and manipulation.
  • Collaborate with AI researchers and robotics engineers to bring research into real-world applications.
  • Help set technical direction and research priorities for advanced AI systems.
  • Work remotely with a US-based research team from approved locations.

Related Jobs