Senior AI Engineer ( Annotation & VLA Models)

CNTXT AI is seeking a Senior AI Engineer to lead the VLA and annotation work for their robot learning systems team in Abu Dhabi, United Arab Emirates. This role involves selecting, fine-tuning, and improving VLA models, owning the annotation of robot data, and contributing to cutting-edge robotics research.

CNTXT AI is a data-first AI company focused on enabling organizations to prepare, build, deploy, and scale sovereign AI solutions with complete data control. They offer enterprise services including high-quality training data annotation and AI solutions that maintain data sovereignty.

If you are passionate about advancing robotics and have a strong background in deep learning and VLA models, we encourage you to apply.

Company Overview

CNTXT AI is a data-first AI company that helps organizations prepare, build, deploy, and scale sovereign AI solutions with complete data control. Their enterprise services include high-quality training data through labeling and annotation for AI labs, enterprise data teams, and robotics applications, alongside AI solutions that integrate with clients' existing systems while keeping data fully sovereign. Their proprietary AI product portfolio includes Munsit, the most accurate Arabic voice AI.

From raw data to production-ready AI, they help organizations and governments move faster without sacrificing compliance, sovereignty, or real-world performance.

Quick Details

• Salary Range: Not Specified
• Job Type: Full-time
• Experience: 5 Year+
• Qualifications: bachelors-degree

Key Responsibilities

• Lead the VLA and annotation work of the team, selecting, fine-tuning, and improving VLA models.
• Own the annotation of robot data, focusing on label quality and policy improvement.
• Utilize published VLA research to develop policies for robot tasks and compare results with published experiments.
• Design and test new methods for post-training, data efficiency, and inference speed.
• Define evaluation protocols for simulation and real robots, analyzing failures and performance.
• Ensure policies are fast enough for real-time control using techniques like TensorRT and mixed precision.
• Measure demonstration requirements and test methods for data efficiency.
• Test new research ideas on robots and write technical reports on experiments.
• Manage annotation QA for robot data and define label schemas for VLA training.
• Use VLA experiment results to guide future labeling and data capture efforts.

Requirements

• 5+ years of experience with deep learning and 3+ years with robot policies or embodied AI.
• Experience fine-tuning VLA models or imitation policies and operating them on real robots.
• Strong knowledge of robot imitation learning (behavior cloning, DAgger) and good knowledge of reinforcement learning for robots.
• Expert knowledge of PyTorch, including fine-tuning large transformer models on multi-GPU systems.
• Good knowledge of an open-source code base for VLA models (e.g., LeRobot, openpi, Isaac-GR00T).
• Experience with large multimodal or video datasets for model training, including data pipelines and versioning.
• Experience designing evaluations for robot policies with fixed test sets and confidence intervals.
• Experience with fast inference for real-time control using tools like TensorRT.
• Evidence of senior work (e.g., conference papers, open-source code, production systems).
• Experience clearly explaining results and risks to engineers and managers.

Benefits

• Opportunity to lead VLA and annotation work in a cutting-edge robotics team.
• Work with advanced VLA models and contribute to novel robotics research.
• Gain experience with a variety of robotics platforms and simulation tools.
• Collaborate with a small, senior technical team focused on pushing the state-of-the-art.
• Contribute to the development of sovereign AI solutions with complete data control.
• Potential for team leadership as the team grows.