PinnedMember-onlyEverything I Hate About Machine LearningI recently gave a talk about Machine Learning in Production, and had a really great time making this slide: Throughout the years I’ve worked with ML, I’ve very quickly become an ML pessimist. …Machine Learning4 min readMachine Learning4 min read
Published in Towards Data Science·PinnedMember-onlyComponents of a Production ML System Using Only PythonA foundational base to learn—and build on—the basics of MLOps components — I recently gave a talk on Machine Learning in Production for NYC Data Science Academy from which I’ve adapted this post. The purpose of this article is to demonstrate the basic functionality of some components of a production ML system using only Python code. …Machine Learning6 min readMachine Learning6 min read
Published in Towards Data Science·PinnedMember-onlyThe Rise of the Term “MLOps”Properly Operationalized Machine Learning is the New Holy Grail — “MLOps (a compound of Machine Learning and “information technology OPerationS”) is [a] new discipline/focus/practice for collaboration and communication between data scientists and information technology (IT) professionals while automating and productizing machine learning algorithms.” — Nisha Talagala (2018) For folks interested, I’ll also be teaching a Machine Learning in Production class…Data Science6 min readData Science6 min read
Published in Towards Data Science·Dec 23, 2022Member-onlyBuilding a Lil’ Stateful, ML Application for Online LearningUsing River, Flask, Gunicorn and Multiprocessing to Build an Application for Online Learning — Most real-time ML systems that I see today are stateless — they train and serve a fixed model artifact for until being completely replaced by another artifact trained on a window of more recent data. …Machine Learning7 min readMachine Learning7 min read
Published in AWS in Plain English·Nov 11, 2021Member-onlyMachine Learning in Production at the AWS re:Invent Conference 20215 production-grade Machine Learning talks at the re:Invent conference of 2021 that I’m excited about. — This year’s AWS re:Invent conference is coming, and it should come as no surprise to anyone that the amount of machine learning (ML) content is downright gratuitous.AWS5 min readAWS5 min read
Published in Towards Data Science·Aug 7, 2021Member-onlyKubernetes Practical Intro for Data ScientistsYup, it’s yet another “intro to k8s” article. The world didn’t need it, you didn’t ask for it, and yet I already had it in a gist somewhere so I thought: “why not”? — I presented this tutorial live recently for New York City Data Science Academy, it was recorded here. I’ve adapted the slides and gist to this single post for convenience. For folks interested, I’ll also be teaching a Machine Learning in Production class in 2022–23!Kubernetes6 min readKubernetes6 min read
Published in Towards Data Science·Jul 17, 2021Member-only5 Different Ways to Deploy your Machine Learning Model with AWSOn the pros and cons of different approaches to getting your model out into the world. — As a data science mentor, I get a lot of questions about the infrastructure side of machine learning. Many newcomers build models they want to expose to the internet via an API, but even with the wealth of available resources they still have trouble doing so. …AWS6 min readAWS6 min read
Published in Towards Data Science·Mar 2, 2021Member-onlyThe Fastest Way to Deploy Your ML App on AWS with Zero Best PracticesPublish a Machine Learning API to the Internet in like…15 Minutes — You’ve been working on your ML app and a live demo is coming up fast. You wanted to push it to Github, add Docker, and refactor the code but you spent all day yesterday on some stupid pickle error you still don’t really understand. Now you only have 1 hour…AWS5 min readAWS5 min read
Published in Towards Data Science·Jan 7, 2021Member-onlyExposing Tensorflow Serving’s gRPC Endpoints on Amazon EKSgRPC with kubernetes, nginx and tensorflow serving — Tensorflow serving is popular way to package and deploy models trained in the tensorflow framework for real time inference. Using the official docker image and a trained model, you can almost instantaneously spin up a container exposing REST and gRPC endpoints to make predictions. Most of the examples and documentation…TensorFlow9 min readTensorFlow9 min read
Published in Towards Data Science·Nov 17, 2020Member-onlyA Simple MLOps Pipeline on Your Local MachineDocker Desktop, Kubernetes and Other Open Source Tooling to Demonstrate Machine Learning Process on your Laptop — It can be difficult to get any kind of ops experience outside of enterprise without turning to the cloud. The simple reason is that there’s little benefit in not turning to the cloud. In order to provide a zero-cost entry point into MLOps for those new to the space, I…Mlops14 min readMlops14 min read
