AI Research Engineer – Multi-Modal, Vision

This is a fully remote position, open to applicants in United Arab Emirates (UAE).

📋 Description

• Conduct comprehensive research and engineering on vision-language models, encompassing training, evaluation, and optimization throughout the entire model development lifecycle.

• Design and execute post-training pipelines that include supervised fine-tuning, knowledge distillation, and reinforcement learning guided by human feedback.

• Develop and sustain high-quality multimodal datasets, involving data curation, filtering, and balancing tailored for specific domain tasks.

• Enhance model efficiency and deployability by adapting models for environments with limited resources through compression and optimization techniques.

• Create and implement evaluation frameworks and benchmarks aimed at assessing model performance, robustness, and success in real-world applications.

• Construct and scale training workflows utilizing distributed GPU infrastructure.

• Identify and address bottlenecks in training pipelines to attain state-of-the-art model quality on designated benchmarks.

• Contribute to and utilize open-source ecosystems, including models, datasets, and tools, to expedite development.

• Remain updated on the latest advancements in multimodal learning and vision-language systems, applying relevant insights to enhance practical applications.

• Publish research outcomes in leading AI conferences and journals whenever applicable.

⛳️ Requirements

• A degree in Computer Science, Machine Learning, or a related discipline; an MS/PhD is preferred.

• Extensive experience with multimodal post-training workflows, including supervised fine-tuning, knowledge distillation, and reinforcement learning based on feedback.

• Practical experience with parameter-efficient fine-tuning and distributed training frameworks.

• Proven capability to develop and refine vision-language models, demonstrating measurable success on standard benchmarks or real-world tasks.

• Experience in adapting models for environments with limited resources.

• Established open-source contributions in multimodal AI on platforms like GitHub or HuggingFace.

• Publications in top-tier AI conferences such as NeurIPS, ICML, ICLR, CVPR, ECCV, etc.

🏝️ Benefits

• Health insurance

• Flexible work arrangements

• Professional development opportunities