The Institute for Ethical AI & Machine Learning

Kompute joins the Linux foundation @LFAIDataFdn to further the cross-vendor GPU Acceleration ecosystem for AI, Machine Learning and Advanced Data Processing use-cases 🚀

Landed in London for #KubeCon Europe! If you are around let's catch up! Also join us tomorrow in our opening keynote at the KubeCon AI Day on the state of GenAI & ML in the Cloud Native Ecosystem 🚀

🥳🥳🥳 Happy New Year 2025 🥳🥳🥳 How time flies! We are celebrating 6 years since starting this weekly MLE newsletter 🎉🎉🎉 Check out the deep dives and resources in this week's edition: buff.ly/3U7DCrp

Insightful news, resources and frameworks in the Machine Learning Ecosystem this week: National AI Strategies & Regulation, Netflix's Large Scale Time Series Infra, Gauge on GenAI's Tech Debt Costs, Gradient Flow on AI Engineers, MLOps Organisational Setups + more 🚀 Check out the deep dives and resources in this week's edition! For anyone looking for exciting ways to develop your ML Engineering skills in 2024, you can join 60,000+ ML practitioners & enthusiasts for weekly news, tutorials articles and MLOps events 📅 + more 🚀 #ML #MachineLearning #ArtificialIntelligence #AI #MLOps #AIOps #DataOps #augmentedintelligence #deeplearning #privacy #kubernetes #datascience #python #bigdata

Central functions support machine learning ambitions across organisations; only ~30% of organizations have a central Data / ML platform, and less than 10% have central AI Risk Governance: Establishing central functions to support teams is a growing trend in machine learning maturity across organisations, an we see some interesting trends, with less than 10% or organisations with a central "AI Inventory", or an "AI Risk & Governance Function"; less than 15% with a central dev productivity function, and; slightly over 30% having a central Data Platform or central ML Platform 💻 We are uncovering great insights as part of our survey on The State of Production ML in 2024; please contribute to this valuable investigation on machine learning tools and platforms used in your production ML development. Your input will help create a comprehensive overview of common practices, tooling preferences, and challenges faced when deploying models to production, ultimately benefiting the entire ML community 🚀 Survey: bit.ly/state-of-ml-20… Results: buff.ly/3YE2rN7 -- If you liked this article you can join 60,000+ practitioners for weekly tutorials, resources, OSS frameworks, and MLOps events across the machine learning ecosystem: buff.ly/3U7DCrp #ML #MachineLearning #ArtificialIntelligence #AI #MLOps #AIOps #DataOps #augmentedintelligence #deeplearning #privacy #kubernetes #datascience #python #bigdata

We are pleased to announce that our next speaker is Alejandro Saucedo on November 18. Register and find out more! ieai.sot.tum.de/event/speaker-… @chluetge @AxSaucedo #ieai.sot.tum.de #ai #artificialintelligence #aiethics

In Machine Learning, Spreadsheets just keep coming back - this time as one of the top choices for experiment tracking and model registry tools... Only a week since we launched our survey on the State of Production ML we are already seeing great insights, help us map the broader ML ecosystem together! We have designed the questions to provide meaningful insights on the current landscape of production ML in 2024 - if you have a chance we would be grateful if you could spend a few minutes on the survey, as you'll contribute valuable information about the machine learning tools and platforms you use in your production ML development. Your input will help create a comprehensive overview of common practices, tooling preferences, and challenges faced when deploying models to production, ultimately benefiting the entire ML community 🚀 Survey: bit.ly/state-of-ml-20… -- If you liked this article you can join 60,000+ practitioners for weekly tutorials, resources, OSS frameworks, and MLOps events across the machine learning ecosystem: buff.ly/4813RDa #ML #MachineLearning #ArtificialIntelligence #AI #MLOps #AIOps #DataOps #augmentedintelligence #deeplearning #privacy #kubernetes #datascience #python #bigdata

Fantastic episode of"Talk Python Podcast" on"Python at Netflix" featured two guests from Netflix's Python Infrastructure team, Amjith Ramanujam and Srinivasan Ramanujam 📺 youtube.com/watch?v=yPBslO… #ML #MachineLearning #ArtificialIntelligence #AI #MLOps #AIOps #DataOps

Meta / Facebook has released their "approach to explaining (ML) ranking" 💡 This includes how posts are ranked in newsfeed, recommendations, ranking of comments, friend recommendations, notifications and more. transparency.fb.com/features/expla… #ML #MachineLearning

New free online course on "Large Language Models: Application through Production" 💡 A great resource for developers, data scientists, and engineers who aim to build applications powered by large language models. edx.org/course/large-l… #ML #MachineLearning

Future of Custom & Fine-Tuned LLMs 🔮 A great article from Outerbounds, the company behind Metaflow, which explores instruction tuning for large language models. outerbounds.com/blog/custom-ll… #ML #MachineLearning #ArtificialIntelligence #AI #MLOps #AIOps #DataOps

ML System Design Case Studies 🖥️ Evidently AI created a fantastic resource containing 200 real-word case studies of ML Systems from 64 companies including Netflix, Airbnb, and Doordash. evidentlyai.com/ml-system-desi… #ML #MachineLearning #ArtificialIntelligence #AI #MLOps #AIOps

🚀 This week in Machine Learning at Scale 🚀 In this edition we cover: 1) ML System Design Case Studies 2) Future of Custom Tuned LLMs 3) LLM App to Prod Databaricks 4) + more 🚀 linkedin.com/pulse/issue-23… #ML #MachineLearning #ArtificialIntelligence #AI #MLOps #AIOps #DataOps

Dealing with Train-serve Skew in Real-time ML Models: A Short Guide 🔭 A comprehensive article that demystifies the training-serving skew on ML models, which arises due to differences between training and serving environments in real-time ML models building.nubank.com.br/dealing-with-t… #ML

Stanford has put together an informative repo which introduces the "Foundation Model Transparency Index"; a framework which rates major AI providers like OpenAI and Google on their adherence to requirements specified in the EU Parliament's AI Act. Repo: github.com/stanford-crfm/…

Resources: Repo: github.com/jacopotagliabu… Paper: arxiv.org/abs/2104.09423

4) Great Expectations for data quality (Alternatives: dbt-expectations plugin) 5) Weights&Biases for experiment tracking (Alternatives: Comet, Neptune) 6) Sagemaker / Lambda for model serving.

Building MLOps at Reasonable Scale: You Don't Need a Bigger Boat ⛵ This resource addresses the challenges of implementing recommender systems at a "reasonable scale" with a case study in the retail industry 🧵 👇

New course from Andrew Ng's DeepLearning.AI in collaboration with AWS on Generative AI 🚀 It includes data gathering, model selection, performance evaluation, deployment, and how to adapt and optimize models to various use cases. deeplearning.ai/courses/genera… #ML #AI