AI Models MAD - Model Automation and Dashboarding

Advanced Micro Devices, Inc•Santa Clara, CA

14d

About The Position

At AMD, our mission is to build great products that accelerate next-generation computing experiences—from AI and data centers, to PCs, gaming and embedded systems. Grounded in a culture of innovation and collaboration, we believe real progress comes from bold ideas, human ingenuity and a shared passion to create something extraordinary. When you join AMD, you’ll discover the real differentiator is our culture. We push the limits of innovation to solve the world’s most important challenges—striving for execution excellence, while being direct, humble, collaborative, and inclusive of diverse perspectives. Join us as we shape the future of AI and beyond. Together, we advance your career. THE ROLE: AMD is looking for a skilled and motivated software engineer to join the Model Automation and Dashboarding (Framework MAD) team — a group focused on ensuring the reliability, performance, and scalability of AI models running on AMD hardware. As part of this team, you’ll build and maintain tools and infrastructure that automate functional and performance validation of deep learning models across ROCm and GPU platforms. Your contributions will directly impact developer confidence, model portability, and transparent benchmarking for internal teams and the open-source community. THE PERSON: We are seeking a software developer with strong technical expertise in Python and Linux-based systems, who is passionate about quality assurance, benchmarking, and automation in the AI/ML space. The ideal candidate thrives in both collaborative and independent environments, demonstrates excellent problem-solving skills, and takes ownership in defining goals and delivering impactful solutions. Experience working with machine learning frameworks, performance dashboards, or automation platforms is a strong plus.

Requirements

Programming & Tooling: Strong Python development skills, with experience in test automation, CI/CD, and Linux scripting.
Machine Learning Workflow Understanding: Familiarity with AI frameworks (e.g., PyTorch, TensorFlow), model benchmarking, and ML model lifecycles.
Performance Analysis: Strong experience with profiling tools, system monitoring, or regression tracking systems for deep learning models.
DevOps & Dashboards: Solid experience in performance dashboards, visualization tools (e.g., Grafana, Plotly), and metrics collection pipelines.
Software Engineering Practices: Proficiency with version control (GitHub), testing strategies, code reviews, and collaborative software development.
Communication & Ownership: Strong written and verbal communication skills with a proactive approach to defining and driving development efforts.

Responsibilities

Model Testing & Validation: Automate functional and performance testing of AI models across ROCm-supported hardware using scalable tools and pipelines.
Software Engineering Excellence: Proficiency in Python and C++ with deep experience in performance tuning, debugging, and robust test design, ensuring reliable, maintainable, high-performance codebases.
Benchmarking Infrastructure: Develop tools for continuous benchmarking and regression tracking across hardware generations and ROCm releases.
Dashboard & Metrics Development: Build and maintain real-time dashboards that report relevant performance, accuracy, and reliability metrics for both internal and public users.
Ecosystem Integration: Collaborate with teams like Deep Learning Models (DLM) and MADengine to support a wide range of models, including public and private/NDA workloads.
Client Enablement: Ensure out-of-box confidence for ROCm clients by validating model performance and functionality in standardized and reproducible environments.
Scalable Tooling: Contribute to the design of portable, easy-to-use Python interfaces that support multi-node profiling, distributed workloads, and containerized deployments.
Open-Source Contributions: Support public-facing MAD GitHub repositories and Docker releases, enabling the community to run and validate models on ROCm.