The evolution of Machine Learning Infrastructure

A curated list of resource links for evolution of Machine Learning Infrastructure

• The rise of Machine learning Infrastructure
• Evoluation 1 : Foundation Models as a service
• Evoluation 2 : Data Centric AI
• Evoluation 3 : Open MLOps
• Books recommendations

The rise of Machine Learning Infrastrucure

• Hidden Technical Debt in Machine Learning Systems - Google
• Infrastructure 3.0 for AI revolution - Venturebeat
• Michelangelo: Uber’s Machine Learning Platform - Uber
• Bighead: Airbnb’s End-to-End Machine Learning Infrastructure - Airbnb

Evolution 1 : Foundation Models as a service

• AI Software Market prediction to 2030 - ARK Investment Management LLC, 2022
• On the Opportunities and Risks of Foundation Models - Stanford University
• The Center for Research on Foundation Models (CRFM) - Stanford University
• ChatGPT — Conversational AI - Open AI
• DALL-E — Text-to-image generation - Open AI
• Stable Diffusion — Text-to-Image generation - Stability AI

Evolution 2 : Data Centric AI

• What is Data Centric AI ? - Data-centric AI Resource Hub
• Why Does Data-Centric AI Matter? - Landing ai
• Data quality and artificial intelligence – mitigating bias and error to protect fundamental rights - FRA
• Training language models to follow instructions with human feedback - Open AI
• What Is a Feature Platform for Machine Learning? - Tecton
• Vertex AI Feature Store - Google
• Feature Stores for ML - Featurestore Org
• Evaluation Stores: Closing the ML Data Flywheel? - Moussa Taifi PhD
• Inside Meta's AI optimization platform for engineers across the company - Meta
• What is ML Observability? - Arize

Evolution 3 : Open MLOps

• Machine Learning Operations (MLOps):Overview, Definition, and Architecture - KIT - Karlsruher Institut für Technologie
• CS 329S: Machine Learning Systems Design - Stanford University
• Chronon - Airbnb
• Jupyter - The Jupyter OSS
• Metaflow - Netflix
• MLflow - Databricks
• Ray - Anyscale
• kubeflow - Kubeflow OSS
• Alibi - Seldon
• Full-Spectrum ML Model Monitoring at Lyft - Lyft

Books recommendations

• Design Machine Learning Systems - A gem of 2022 in Machine Learning System Design. It will introduce you to the entire Machine Learning Lifecycle and prepare you for further deep dives
• Accelerate - ML and MLOps Engineers are meant to bring Software Engineering practices to the Data Science world. After reading this book you will understand DevOps practices in and out
• Machine Learning Design Patterns - The book introduces you to 30 Design Patterns for Machine Learning. You will find 30 recurring real life problems in ML Systems, how a given pattern tries to solve them and what are the alternatives. Always have this book by your side and refer to it once you run into described problems - the book is gold
• Team Topologies - As ML or MLOps Engineer you might be placed into different types of teams. It could be ML Platform Team, Stream Aligned cross-functional Team or an Enabling Team. After reading this book - you will understand the purpose of these different Team Types and what the most efficient communication patterns between them are
• Infrastructure as Code - This is the area that I see under-looked most when it comes to ML and MLOps Engineers. Most of the time you would be working in the cloud and your day-to-day would include a lot of IaaC, especially if you are part of a ML Platform Team. Learning IaaC will give you the edge in today's competitive markets