# model-training Published articles for model-training. This is one page of public article previews, not the complete archive. Follow Next page to continue. Summaries are not the original full articles. ## Blue Proton Initiative: four 17-year-olds are building AI-powered livestock monitoring with Arduino DevFeed: [Blue Proton Initiative: four 17-year-olds are building AI-powered livestock monitoring with Arduino]() Original publisher: [Read original article]() Author: Arduino Team Published: 2026-08-27T13:18:53Z Content type: article Language: en Sources: [Arduino Blog]() Topics: [Arduino](), [Artificial Intelligence](), [Computer vision](), [Neural Network](), [object-detection](), [C](), [datasets]() Tags: [ai](), [ai-powered-livestock-monitoring](), [arduino](), [computer-vision](), [data](), [livestock-monitoring](), [model-training](), [monitoring](), [neural-networks](), [uno-q](), [yolo]() ### AI overview An Arduino Blog article profiles four 17-year-olds in Italy developing an AI-powered livestock monitoring system with the UNO Q. The project uses computer vision, custom-trained neural networks, and diverse image data to identify animals, count them, and detect possible health problems in real time. ### Source excerpt Pietro Maria Piazza, Alessandro Nesci, Davide Santucci, and Matteo Angiolillo are not waiting to finish school before starting to build something real. Based in Forlì, Italy, the four friends behind Blue Proton Initiative strive to develop an AI-powered livestock monitoring system designed to help farmers identify individual animals and detect early signs of health problems [...] The post Blue Proton Initiative: four 17-year-olds are building AI-powered livestock monitoring with Arduino appeared first on Arduino Blog. ## How Adobe Reduced GPU Idle Time in Generative AI Training Through Faster Data Access and Checkpointing DevFeed: [How Adobe Reduced GPU Idle Time in Generative AI Training Through Faster Data Access and Checkpointing]() Original publisher: [Read original article]() Author: Maddie Presland Published: 2026-08-25T15:22:30Z Content type: article Language: en Sources: [Backblaze Blog | Cloud Storage & Cloud Backup]() Topics: [GPU](), [Training AI Models](), [cloud-infrastructure](), [networking](), [data]() Tags: [ai-ml](), [ai-training](), [b2cloud](), [cloud](), [cloud-storage](), [compute](), [ethernet](), [featured](), [featured-cloud-storage](), [gpu](), [model-training](), [networking]() ### AI overview Adobe's generative AI training pipeline left roughly two-thirds of GPU time waiting for data. The article attributes the waste to storage and retrieval bottlenecks, networking limits, and checkpointing overhead, and describes Adobe's use of a high-performance networking fabric and fragmented checkpoint storage to reduce delays. ### Source excerpt Adobe's experience reveals why GPUs sit idle during AI model training: slow storage, insufficient throughput, and uneven workloads. Learn how storage bottlenecks and uneven data loading leave expensive GPUs idle--and how always-hot, high-throughput object storage keeps AI training pipelines running efficiently at scale while reducing wasted compute costs. The post Why Your GPU Is Sitting Idle: The Data Pipeline Problem No One Talks About appeared first on Backblaze Blog | Cloud Storage & Cloud Backup ## Agent Night demo recap: Evil Martians on making your docs discoverable to agents DevFeed: [Agent Night demo recap: Evil Martians on making your docs discoverable to agents]() Original publisher: [Read original article]() Author: WorkOS Published: 2026-08-17T19:04:06Z Content type: article Language: en Sources: [WorkOS Blog]() Topics: [GitHub](), [dev-tools](), [Large Language Model](), [Search engine optimization (SEO)]() Tags: [agent](), [agents](), [developer-tools](), [docs](), [github](), [llm](), [model-training](), [retrieval]() ### AI overview Irina Nazarova's Agent Night demo recap explains how developer-tool documentation can be discovered by agents through websites and GitHub repositories. It distinguishes retrieval, which depends on access to documentation at request time, from training, which depends on whether content enters model corpora, and emphasizes repository descriptions, topics, licensing, and Markdown access. ### Source excerpt Irina Nazarova's Agent Night lightning demo: why GitHub, not your docs site, is the path into model training sets, and how Evil Martians measured all of it. ## How Keras 3 Helped Modernise Expedia Group's Lodging Ranking Stack DevFeed: [How Keras 3 Helped Modernise Expedia Group's Lodging Ranking Stack]() Original publisher: [Read original article]() Author: Conor Worthington Published: 2026-08-11T11:01:02Z Content type: article Language: en Sources: [Expedia]() Topics: [Keras](), [recommendation systems](), [Latency](), [Inference Performance](), [Machine learning]() Tags: [data-science](), [features](), [framework](), [improvements](), [inference](), [keras](), [latency](), [llms](), [machine-learning](), [mixture-of-experts](), [model-training](), [neural-networks](), [ranking](), [recommendation-system]() ### AI overview Expedia Group describes modernising its lodging-ranking stack around Keras 3. The changes included rewriting parts of its pipelines, making model training 30% faster and reducing P99 inference latency by two-thirds. The article discusses limitations of the previous Keras 2, eager-training and combined-serving setup, along with the use of newer APIs and accelerator-oriented optimisations. ### Source excerpt Expedia Group Technology -- DataWhat happened when we treated a framework migration as an architecture modernisation -- and cut P99 inference latency by two-thirdsSt Paul's and millennium bridge, London Expedia Group™ has always been a market leader in providing personalised search experiences for travellers. As our ranking models evolved, we saw an opportunity not just to migrate to Keras 3, but to modernise the broader stack around it so we can better serve travellers. This led us to rewrite key parts of our pipelines that made model training 30% faster and cut P99 inference latency by two-thirds. Our main focus in this blog is to discuss our improvements to lodging ranking -- our service which handles users' search requests and returns a personalised property ranking on top of a lightweight candidate generator. This service is frequently retrained to improve customer experience as we get new signals over time. Example of personalised lodging ranking on Expedia search result page for a London search As such, our velocity to make changes needs to be fast, but more importantly we need to be able to build models which are state-of-the-art, enabling customers to easily find and book the most relevant property for their trip. The problem with Keras 2, eager training and combined serving The challenge was not that the old stack was broken. It was that it had gradually become a limiting factor. On the training side, we were carrying dependencies on older components and missing out on modern Keras APIs, new optimisers, new layers and cleaner support for accelerator-oriented training. On the serving side, we had a path that was simple and operationally familiar, but not especially well suited to aggressively optimised inference for heavier ranking architectures. This blockage on utilising new layers and optimisers has become more pronounced as LLMs see blistering amounts of innovation. Naturally, these new features are only available in Keras 3 or as separate backends from ou ## KubeCon + CloudNativeCon India 2026: Our Kubeflow Community Experience DevFeed: [KubeCon + CloudNativeCon India 2026: Our Kubeflow Community Experience]() Original publisher: [Read original article]() Author: Khushi Agrawal Published: 2026-07-27T05:00:00Z Content type: article Language: en Sources: [Kubeflow]() Topics: [Kubernetes](), [MLOps](), [distributed-training](), [model-serving](), [Large Language Model](), [Retrieval Augmented Generation (RAG)](), [GPU](), [Cloud]() Tags: [2026](), [ai](), [cloud-native](), [community](), [distributed-training](), [gpu](), [india](), [kubecon](), [kubernetes](), [llms](), [machine-learning](), [mlops](), [model-serving](), [model-training](), [rag]() ### AI overview A Kubeflow community recap of KubeCon + CloudNativeCon India 2026 in Mumbai. It describes Kubeflow's presence at the CNCF Project Pavilion, common architecture questions, featured talks, booth discussions about production machine learning workflows, distributed training, model serving, GPU resource allocation, LLMs, and RAG pipelines, and efforts to onboard contributors. ### Source excerpt Introduction ## Training Orchestrator: Unifying Model Training at Yelp DevFeed: [Training Orchestrator: Unifying Model Training at Yelp]() Original publisher: [Read original article]() Author: Ying Wang and Nathan Sponberg, Software Engineer Published: 2026-07-14T00:00:00Z Content type: article Language: en Sources: [Yelp]() Topics: [Machine Learning & Artificial Intelligence](), [Orchestration](), [Apache Spark](), [Training AI Models](), [Model Development](), [configuration](), [configuration-management](), [Monitoring]() Tags: [configuration](), [machine-learning](), [maintenance](), [mlflow](), [model-training](), [orchestration](), [reproducibility](), [spark](), [tooling]() ### AI overview Yelp's Core Machine Learning Team developed Training Orchestrator to standardize how machine learning teams define and run Spark-based model training. The configuration-driven system addresses duplicated code, inconsistent configurations, limited local testing, scattered validation and monitoring, and poor reproducibility across environments. ### Source excerpt At Yelp, we train many machine learning models on different schedules. Applied machine learning teams all have their own set of Spark-based training batches, scripts, and configurations. Over time, these diverged, leading to duplicated code, subtle inconsistencies, and a growing maintenance burden. Yelp's Core Machine Learning Team has developed excellent tooling across our ML ecosystem over the years: feature stores for reproducible data, a unified training library for neural networks and gradient-boosted trees, seamless Spark integration, and MLflow services for model tracking and deployment. But there was still one key piece missing right in the middle: a standardized way to... ## Gesture Recognition Based on TFLite DevFeed: [Gesture Recognition Based on TFLite]() Original publisher: [Read original article]() Author: John Lee Published: 2026-04-10T00:00:00Z Content type: tutorial Language: en Sources: [Blog on Developer Portal]() Topics: [TensorFlow Lite](), [Espressif](), [model-deployment](), [Tensorflow](), [Keras](), [On-device AI]() Tags: [ai](), [blog](), [deep-learning](), [deployment](), [development-board](), [edge-ai](), [edge-computing](), [esp-idf](), [espressif](), [how-to](), [inference](), [keras](), [model-deployment](), [model-training](), [tensorflow-lite]() ### AI overview This tutorial demonstrates gesture recognition on Espressif SoCs using TensorFlow Lite Micro. It covers data collection, model training, conversion for TFLite Micro, and deployment with C++ code for model loading, preprocessing, and inference. ### Source excerpt This article demonstrates how to implement gesture recognition using TensorFlow Lite Micro on Espressif SoCs. It covers the complete workflow from data collection and model training to model deployment, showcasing TensorFlow Lite Micro's applications in edge AI. ## Improving Embedding-Based Candidate Generation for Recommender Systems with a Two-Tower Model DevFeed: [Improving Embedding-Based Candidate Generation for Recommender Systems with a Two-Tower Model]() Original publisher: [Read original article]() Author: Ahmad Khan Published: 2026-03-18T21:04:58Z Content type: article Language: en Sources: [Glassdoor Engineering]() Topics: [recommendation systems](), [Machine learning](), [Embeddings](), [Neural Network](), [data](), [ordering]() Tags: [ai](), [data](), [embedding](), [machine-learning](), [model-training](), [neural-network](), [ordering](), [ranking](), [recommender-systems](), [self-supervised-learning](), [two-tower-model]() ### AI overview Glassdoor describes a two-tower neural network for candidate generation in its Community feed recommendation system. The model learns user and post embeddings from interaction data to retrieve potentially relevant posts before filtering, ranking, and ordering. ### Source excerpt Overview At Glassdoor (part of Indeed), millions of professionals come to learn about workplaces, ask career questions, and share experiences through our Community platform. When users open the Community Home Feed, they expect to see posts that are relevant to their interests, industry, and career stage. However, generating that feed is not a simple task. At any moment, thousands of new posts across hundreds of community feeds compete for attention. From discussions about interview experiences to workplace culture questions and industry advice, the platform contains a diverse and constantly evolving stream of content. The challenge is to quickly identify which of these posts are most relevant for each user. To address this problem, we built a multi-stage machine learning recommendation system that efficiently sources, filters, and ranks posts before they appear in the user's feed. In a previous post, Inside Glassdoor's Multi-Stage Recommendation System, we described the architecture of this system and how the different stages work together to deliver personalized content. Snapshot of the Glassdoor Community Home Feed At a high level, the system consists of four stages: Candidate Generation -- retrieves a large set of potentially relevant posts Filtering -- removes posts that are undesirable, duplicated, or already seen Ranking -- predicts which posts a user is most likely to engage with Ordering -- applies heuristics to maintain diversity and freshness in the feed Each stage progressively narrows the set of posts until only the most relevant posts remain. 4 Stage ML RecSys Design In this blog post, we focus on the Candidate Generation stage, which is responsible for retrieving an initial pool of potentially relevant posts from a very large pool of posts across many different feeds. For Candidate Generation, we developed a Two-Tower Neural Network Model that learns embeddings for users and posts based on real interaction data. In the rest of this blog post, we will walk ## How mpathic built better ML workflows by switching from Elasticsearch to ClickHouse Cloud DevFeed: [How mpathic built better ML workflows by switching from Elasticsearch to ClickHouse Cloud]() Original publisher: [Read original article]() Author: ClickHouse Published: 2025-10-10T16:20:55Z Content type: article Language: en Sources: [ClickHouse Blog]() Topics: [clickhouse](), [elasticsearch](), [AI Development](), [Cloud](), [Inference](), [Amazon EC2]() Tags: [2025](), [analytics](), [audio](), [build-faster](), [clickhouse](), [cloud](), [compliance](), [elasticsearch](), [healthcare](), [ml](), [model-training](), [transcription](), [workflows]() ### AI overview mpathic uses AI to analyze therapy-session audio for patient safety and compliance in clinical drug trials. The article explains how noisy recordings require speaker segmentation, fingerprinting, transcription, classification, model training, and analytics, and describes the company's move from Elasticsearch and EC2-based pipelines to ClickHouse Cloud for faster, leaner ML workflows. ### Source excerpt Learn why Elasticsearch was holding mpathic back, and how switching to ClickHouse Cloud helped them build faster, leaner ML workflows. ## Logging and registering models with MLflow DevFeed: [Logging and registering models with MLflow]() Original publisher: [Read original article]() Author: Maria Vechtomova Published: 2025-07-31T18:49:09Z Content type: tutorial Language: en Sources: [MarvelousMLOps]() Topics: [databricks](), [MLOps](), [Model Development](), [scikit-learn]() Tags: [databricks](), [machine-learning](), [mlops](), [model](), [model-training](), [pipeline]() ### AI overview Lecture 4 in an MLOps with Databricks course explains how to log and register machine learning models with MLflow. It covers the standardized MLflow Model format, supported model flavors, custom PythonModel implementations, and a scikit-learn pipeline example. ### Source excerpt Lecture 4 of MLOps with Databricks course ## (LoRA) Fine-Tuning FLUX.1-dev on Consumer Hardware DevFeed: [(LoRA) Fine-Tuning FLUX.1-dev on Consumer Hardware]() Original publisher: [Read original article]() Author: Derek Liu; Marc Sun; Sayak Paul; merve; Linoy Tsaban Published: 2025-06-19T00:00:00Z Content type: tutorial Language: en Sources: [Hugging Face - Blog]() Topics: [Fine-tuning](), [lora](), [flux](), [quantization](), [Training AI Models](), [GPU]() Tags: [artificial-intelligence](), [dev](), [diffusers](), [fine-tuning](), [flux](), [gpu](), [guide](), [hardware](), [inference](), [lora](), [model-training](), [nvidia-rtx](), [quantization](), [training]() ### AI overview This tutorial explains how to fine-tune the FLUX.1-dev diffusion model efficiently with QLoRA on a single consumer GPU using less than about 10 GB of VRAM. It describes the model components, focuses training on the transformer while keeping the text encoders and VAE frozen, and discusses LoRA, quantization, and FP8 training for memory and speed improvements. ### Source excerpt We're on a journey to advance and democratize artificial intelligence through open source and open science. ## Improving Nextdoor Notification Email Engagement with Generative AI and Rejection Sampling DevFeed: [Improving Nextdoor Notification Email Engagement with Generative AI and Rejection Sampling]() Original publisher: [Read original article]() Author: Jaewon Yang Published: 2023-10-16T17:03:52Z Content type: article Language: en Sources: [Nextdoor]() Topics: [Artificial Intelligence](), [Generative AI](), [Reinforcement learning](), [email]() Tags: [a-b-testing](), [ai](), [building](), [chatgpt](), [email](), [generative-ai](), [metrics](), [model-training](), [notifications](), [platform](), [reinforcement-learning]() ### AI overview Nextdoor describes using Generative AI and rejection sampling to improve Notification email subject lines. The approach uses user engagement feedback to produce more informative subject lines intended to increase email opens, clicks, and sessions. ### Source excerpt Generative AI (Gen AI) has demonstrated proficiency in content generation but does not consistently guarantee user engagement, mainly for two reasons. First, Gen AI generates content without considering user engagement feedback. While the content may be informative and well-written, it does not always translate to increased user engagement such as clicks. Second, Gen AI-produced content often remains generic and may not always provide the specific information that users seek. Nextdoor is the neighborhood network where neighbors, businesses, and public agencies connect with each other. Nextdoor is building innovative solutions to enhance the user engagement with AI-Generated Content (AIGC). This post outlines our approach to improving user engagement through user feedback, specifically focusing on Notification email subject lines. Our solutions employ Rejection sampling [1], a technique used in reinforcement learning, to boost the engagement metrics. We believe our work presents a general framework to drive user engagement with AIGC, particularly when off-the-shelf Generative AI falls short in producing engaging content. To the best of our knowledge, this marks an early milestone in the industry's successful use of AIGC to enhance user engagement. Introduction At Nextdoor, one of the ways to drive user growth and engagement on platform is through emails. One of the emails we have is called New and Trending notifications, where we send a single post that we think the user might be interested in and want to engage with. As part of sending an email, we need to determine a subject line of the email for the email audiences. Historically, we simply pick the first few words of the post being sent to be the subject line. However, in certain posts, these initial words are often greetings or introductory remarks and may not provide valuable information to the user. In the provided image example below, we observe a simple greeting, "Hello!" Figure 1. New and Trending email wher ## Implementing Multi-GPU Distributed Training for Stitch Fix's Personalized Recommendations DevFeed: [Implementing Multi-GPU Distributed Training for Stitch Fix's Personalized Recommendations]() Original publisher: [Read original article]() Published: 2023-06-08T09:00:00Z Content type: article Language: en Sources: [Stitch Fix]() Topics: [distributed-training](), [recommendations](), [GPU](), [sharding](), [Training AI Models](), [Multi-GPU](), [PyTorch]() Tags: [distributed-training](), [gpu](), [model-training](), [multi-gpu](), [parallel](), [performance](), [pytorch](), [recommendations](), [sharding]() ### AI overview This Stitch Fix engineering article explains how the company implemented multi-GPU distributed training for its Client Time Series Model (CTSM), a PyTorch-based model used in personalized recommendations. It describes sharding training data across GPUs and training mini-batches in parallel to reduce training time, along with the surrounding retraining and deployment workflow. ### Source excerpt Stitch Fix uses a cutting-edge multi-tiered recommender system stack to personalize styling recommendations at scale. This stack comprises several critical components, including feature generation, scoring, ranking, and inventory optimization techniques. Our scoring module is based on the Client Time Series Model (CTSM) which is an award winning novel sequence based model that uses temporally masked encoders. CTSM is built using PyTorch, and was initially trained on a single Graphics Processing Unit (GPU) instance. Since we first put this model into production last year, we have launched several updates to the model that improved its performance. Many of these improvements involved adding new features or increasing the time window of our training data. As a result, the model training time increased significantly, making it harder for us to iterate quickly and get feedback on new ideas we want to try for improving the model. We needed a way to reduce the model training time. This blog delves into the steps we followed to overcome this challenge and our journey to implement multi-GPU distributed model training for CTSM. By sharding the training data across multiple GPUs and training multiple mini-batches in parallel, we aimed to achieve significant reductions in training time. We present empirical results showcasing the observed reduction in training time when we scaled up resources from 1 to N GPUs, and share some future directions we are considering in our continued effort to speed up model training. Model Training Workflow The scores generated by CTSM are leveraged by multiple downstream services to get insight into what items a client is likely to purchase. The model is retrained at a regular cadence to ensure that it is using the most updated information about each client when making predictions and does not degrade in its performance. We leverage configuration driven machine learning pipelines to set up a Directed Acyclic Graph (DAG) that automatically retrains ## LinkBERT: Improving Language Model Training with Document Link DevFeed: [LinkBERT: Improving Language Model Training with Document Link]() Original publisher: [Read original article]() Author: A Href; Michihiro Yasunaga; Jure Leskovec; Percy Liang Published: 2022-05-31T07:00:00Z Content type: article Language: en Sources: [The Stanford AI Lab Blog]() Topics: [Large Language Model](), [Natural language processing](), [Training AI Models](), [Self-supervised learning](), [Web](), [data](), [Fine-tuning]() Tags: [data](), [google](), [language-models](), [model](), [model-training](), [natural-language-processing](), [self-supervised-learning](), [training](), [web]() ### AI overview The article introduces LinkBERT in the context of language-model pretraining. It explains how BERT and GPT use self-supervised objectives to learn from large text corpora, then describes a limitation of treating documents independently: hyperlinks and citation links connect knowledge across documents and can support multi-hop understanding. The supplied text ends while introducing this challenge. ### Source excerpt Language Model Pretraining Language models (LMs), like BERT 1 and the GPT series 2, achieve remarkable performance on many natural language processing (NLP) tasks. They are now the foundation of today's NLP systems. 3 These models serve important roles in products and tools that we use every day, such as search engines like Google 4 and personal assistants like Alexa 5. These LMs are powerful because they can be pretrained via self-supervised learning on massive amounts of text data on the web without the need for labels, after which the pretrained models can be quickly adapted to a wide range of new tasks without much task-specific finetuning. For instance, BERT is pretrained to predict randomly masked words in original text (masked language modeling), e.g. predicting the masked word "dog" from "My __ is fetching the ball". GPTs are pretrained to predict the next word given a previous sequence of text (causal language modeling), e.g. predicting the next word "ball" from "My dog is fetching the". In either cases, through pretraining, LMs learn to encode various knowledge from a text corpus that helps to perform downstream applications involving language understanding or generation. In particular, LMs can learn world knowledge (associations between concepts like "dog", "fetch", "ball") from training text where the concepts appear together, and help for knowledge-intensive applications like question answering. 6 Challenges. A challenge with most common LM pretraining strategies is that they model a single document at a time. That is, one would split a text corpus into a list of documents and draw training instances for LMs from each document independently. Treating each document independently may pose limitations because documents often have rich dependencies with each other. For instance, text from the web 7 or scientific literature 8 is often used for LM training, but they all have document links, such as hyperlinks and citation links. Document links are crucial bec ## Lessons Learned From Running Apache Airflow at Scale DevFeed: [Lessons Learned From Running Apache Airflow at Scale]() Original publisher: [Read original article]() Author: Megan Parker Published: 2022-05-23T14:00:01Z Content type: article Language: en Sources: [Shopify Engineering](), [Shopify Engineering - Shopify Engineering]() Topics: [airflow](), [Orchestration](), [Shopify](), [Kubernetes](), [Google Cloud Platform (GCP)](), [MySQL](), [Apache Iceberg](), [Training AI Models](), [data-modeling]() Tags: [airflow](), [apache](), [cloud](), [cloud-storage](), [data-modeling](), [google-cloud](), [kubernetes](), [load](), [machine-learning](), [model-training](), [monitoring](), [mount](), [mysql](), [orchestration](), [scale](), [scheduler](), [shopify](), [solutions](), [workflows]() ### AI overview Shopify explains lessons from operating Apache Airflow at scale for more than two years. The deployment runs on Kubernetes with the Celery executor and MySQL, supporting data extraction, machine learning model training, Apache Iceberg table maintenance, and DBT-powered data modeling. The article describes scaling to more than 10,000 DAGs, over 400 concurrent tasks, and more than 150,000 daily runs, along with challenges involving file access, DAG capabilities, traffic variability, and resource contention. ### Source excerpt Shopify shares some of the lessons we learned and solutions we built in order to run Airflow at scale. ## MLOps in Glassdoor: an à la carte approach DevFeed: [MLOps in Glassdoor: an à la carte approach]() Original publisher: [Read original article]() Author: Zixin Wu Published: 2022-01-12T00:42:43Z Content type: article Language: en Sources: [Glassdoor Engineering]() Topics: [MLOps](), [Machine learning](), [data](), [Software Engineering]() Tags: [airflow](), [data](), [devops](), [machine-learning](), [mlflow](), [mlops](), [model-training](), [monitoring](), [offline](), [orchestration](), [software-engineering]() ### AI overview The article introduces Glassdoor's MLOps approach, explaining how machine-learning systems differ from traditional software because they depend on changing data, model training, data-centric evaluation, version tracking, and monitoring. It presents MLOps as the application of DevOps practices to ML systems through automation and monitoring. ### Source excerpt Authors: Zixin Wu, Srinivasan Ramaraju Introduction Data Science and Machine Learning (ML) have become an essential toolkit for businesses to increase their competitiveness and provide greater value to their customers in recent years, largely thanks to the abundantly available large data sets, more cost-effective computation resources, and advancements in ML algorithms. In Glassdoor, we utilize ML to build intelligent products which help our job seekers find the jobs and companies they love. While ML can unlock new capabilities for businesses, it comes with its complexity and challenges compared to traditional software engineering, mainly due to data dependency and team skillset shift. Data dependency: a major reason for using ML approaches to build complex software is exactly because the desired behavior of the software is so complex that it's more efficient to be driven by data. This dependency brings new components in the software development process: Instead of simply compiling source codes, now the software (ML models) needs to be trained/retrained by (usually a large amount of) data before it can be used to serve clients Evaluation and test of ML model behaviors is data-centric and often involves offline experimentation Version tracking of ML models needs to consider both code and data changes The data could change over time due to various reasons, such as the way the data is collected or processed, or customer behavior changes which, in turn, could be driven by the effect of using the ML model. This essentially creates a hidden feedback loop. All these changes in data need to be monitored and acted accordingly. Team skillset: in many companies, the majority of the ML team members are data scientists or ML researchers, specialized in dealing with data and ML algorithms. However, to integrate an ML solution to existing systems and continuously operate it in production, many more components are needed besides ML model training and evaluation. Automating and stan