Open to opportunities

Michael Liu

@michaelliu1

Message

Senior AI engineer specializing in LLMs, RAG, agentic AI, and production-grade MLOps.

United States

Message

What I'm looking for

I seek roles building and operationalizing production-grade LLM/RAG/agentic AI systems with strong MLOps, cross-functional collaboration, and measurable impact.

I am a Senior Machine Learning Engineer with 9+ years building production AI across consumer SaaS, enterprise platforms, and regulated healthcare environments. I specialize in agentic AI, retrieval-augmented generation (RAG), multimodal systems, and end-to-end MLOps for scalable, reliable deployments.

At Notion and Clarifai I led production-grade RAG and embedding retrieval systems (FAISS), built GPU-accelerated inference services, introduced CI/CD and model versioning, and created observability and evaluation pipelines that reduced incidents and debugging time. I also delivered clinical NLP solutions at Flatiron Health with HIPAA-compliant deployments and human-in-the-loop validation.

I mentor engineers, collaborate cross-functionally with product and client teams, and prioritize safe, grounded AI that meets business needs and ethical guidelines. I focus on measurable improvements in relevance, latency, cost-efficiency, and clinical applicability when moving models from research into production.

Experience

Work history, roles, and key accomplishments

Current

Senior AI Engineer

Current

Notion

Jul 2023 - Present (2 years 7 months)

Led production-grade RAG and agent orchestration systems, improving semantic search and document summarization relevance by ~25% and reducing post-release issues by ~30% through evaluation pipelines and CI/CD for model versioning.

RAG Faiss Python Model Evaluation CI CD For ML Observability

Senior Machine Learning Engineer

Clarifai

Mar 2020 - Jul 2023 (3 years 4 months)

Built GPU-accelerated model serving and embedding retrieval pipelines, improving inference p95 latency by ~15% and retrieval accuracy by ~20% while reducing compute costs ~18–20% through optimization and CI/CD deployments on Kubernetes.

PyTorch TensorFlow Faiss Kubernetes Docker Distributed Systems Apache Spark MLOps Model Versioning

Machine Learning Engineer

Flatiron Health

Sep 2018 - Mar 2020 (1 year 6 months)

Developed HIPAA-compliant clinical NLP pipelines to extract oncology data from unstructured notes, increasing entity extraction precision ~20% and reducing deployment time by ~30% via secure AWS deployments and human-in-the-loop validation.

SpaCy PyTorch AWS HIPAA Human In The Loop Model Evaluation ETL

Software Engineer

Spotify

May 2015 - Sep 2018 (3 years 4 months)

Built backend services and data pipelines for personalization and recommendation systems, improving system error rates by ~15% and enabling real-time recommendations for millions of users through optimized data workflows and A/B testing.

Backend Development Data Pipelines Real Time Recommendations A B Testing Python Performance Optimization