Responsibilities:
Architectural Design
• Design, build, and own the end-to-end lifecycle of machine learning systems—from data
strategy and model architecture to production deployment and monitoring.
• Make principled architectural decisions that balance accuracy, scalability, maintainability, and
cost.
System Optimization
• Diagnose and resolve complex performance issues in model training and inference, including
latency, memory usage, throughput, and convergence behavior.
• Optimize systems across software and hardware boundaries.
Algorithmic Implementation
• Translate high-level business and product requirements into efficient, scalable algorithms
and custom model architectures.
• Implement novel or adapted approaches where off-the-shelf solutions fall short.
Production Engineering
• Bridge the gap between experimental code and reliable, high-availability software services.
• Build reliable services, APIs, and pipelines that support versioning, monitoring, rollback, and
continuous improvement.
Continuous Improvement
• Stay current with the rapidly evolving AI landscape.
• Evaluate new research, tools, and architectural patterns—distinguishing meaningful
structural breakthroughs from short-lived trends.

Technical Requirements:
First-Principles Thinking
• Deep, intuitive understanding of how machine learning models function.
• Ability to debug model behavior and logic—not just syntax or training scripts.
Software Engineering Excellence
• Expert-level programming skills in Python (and/or C++), with a strong emphasis on writing
clean, modular, performance-oriented code.
• Strong understanding of software design, testing, and maintainability.
Modern AI Toolkits
• Hands-on experience with deep learning frameworks (e.g., PyTorch, TensorFlow, JAX).
• Ability to build custom components, training loops, and model architectures when needed.
Data Fluency
• Experience working with large-scale, unstructured datasets.
• Ability to design and implement robust data pipelines that transform raw data into model-
ready inputs.
Hardware Awareness
• Solid understanding of how ML workloads interact with compute resources (GPUs/TPUs).