Machine Learning Engineer — AI Architecture Research

Overview

Our partner company in Saudi Arabia seeks a Machine Learning Engineer specializing in AI architecture research. This role immerses you in pioneering next-gen AI model architectures, transitioning innovative concepts from research to production-ready systems.

Key Responsibilities

• Investigate and create novel neural network architectures, including but not limited to Transformer alternatives, recurrent, hybrid, and long-context models.
• Design and carry out experimental studies on scaling laws, memory mechanisms, training behaviors, and trade-offs involving computation and performance.
• Develop prototypes of models end-to-end, converting theoretical research into practical, training-compatible implementations.
• Analyze model behaviors, recognize failure points, and assess architectural advantages and drawbacks.
• Collaborate closely with systems and inference engineering teams to ensure that new architectures are optimized for efficiency, scalability, and deployment.
• Review, reproduce, benchmark, and extend cutting-edge machine learning research papers and contribute to internal documentation and possibly open-source projects.
• Navigate fluidly between theoretical frameworks, rapid prototyping, and production-focused engineering tasks.

Qualifications

• Deep understanding of machine learning and deep learning principles with hands-on experience in model development.
• Experience implementing neural network architectures from the ground up.
• In-depth knowledge of attention mechanisms, recurrent neural networks (RNNs), state-space models, and hybrid model designs.
• Solid grasp of training dynamics, optimization techniques, scaling phenomena, and performance considerations at the architecture level.
• Insight into constraints related to memory usage, latency, computational efficiency, and deployment feasibility.
• Proficient in PyTorch or JAX frameworks, capable of writing experimental ML research code.
• Skillful in evaluating architectural concepts through theoretical analysis and empirical testing.
• Excellent communication skills to articulate complex technical ideas and architectural trade-offs clearly.
• Preferred experience includes working with non-Transformer architectures such as advanced RNNs, state-space models, or extended context systems.
• Desirable background includes involvement in research-centric startups, contributions to open-source ML projects, experience with large-scale model training, and custom training loops.
• Publications, preprints, significant research contributions, or expertise in inference optimization and deployment considerations are advantageous.

Benefits

• Competitive salary and equity stake in the company.
• Work focused on foundational AI model architecture beyond mere fine-tuning.
• Influential role in steering technical and research directions within a fast-expanding organization.
• Collaborate with a small team of highly skilled professionals fostering rapid feedback within a research-driven culture.
• Engagement from experimental concept stages through to production-level deployment.
• Full-time employment within an internationally distributed working environment.

Additional Information

This position is managed by a partner firm handling the application process and subsequent steps. Applications are reviewed via an AI-supported system for objectivity and speed, but final hiring decisions remain with the company. Data privacy is protected in compliance with applicable regulations.