Lead AWS AI/ML Architect (Remote)

Role Description

This role owns the end-to-end architecture and delivery of AI and ML solutions on AWS. You will design and lead scalable GenAI and ML systems, working closely with product and engineering teams to ship real-world use cases using services like Amazon Bedrock and Amazon SageMaker.

• Own the overall AI/ML architecture on AWS for key products.
• Design and oversee LLM and ML solutions built on Amazon Bedrock and Amazon SageMaker.
• Define and implement ML flows/pipelines: data preparation → training → evaluation → deployment → monitoring.
• Lead and mentor a small team of AI/ML engineers and developers.
• Build and review Python + FastAPI APIs that expose ML and LLM capabilities to internal and external consumers.
• Collaborate with product, data, and engineering stakeholders to shape technical direction and roadmaps.
• Ensure solutions follow AWS best practices for security, cost-efficiency, reliability, and scalability.
• Establish coding standards, code review practices, and best practices for ML engineering and MLOps.

Qualifications

• 5+ years of experience in software or ML engineering, including 1–2 years as a tech lead or architect.
• Strong hands-on experience with AWS AI/ML services, especially:
• Amazon Bedrock
• Amazon SageMaker
• Excellent Python skills and experience building APIs with FastAPI (or similar frameworks).
• Solid understanding of LLM fundamentals: prompts, tokens, context windows, latency/cost trade-offs, and common error modes.
• Strong grasp of ML fundamentals: common algorithms, model evaluation, overfitting/underfitting, and performance metrics.
• Proven experience designing and operating ML flows/pipelines in production environments.
• Demonstrated ability to mentor engineers and collaborate effectively with non-technical stakeholders.

Nice to have:

• Experience with RAG (Retrieval-Augmented Generation) and other GenAI production use cases.
• Familiarity with ML experiment tracking, monitoring, and broader MLOps practices (e.g., feature stores, model registries, CI/CD for ML).