Self-Hosted Databricks Alternative: Homelab Data Lakehouse (AI/ML Data Factory)
Self-Hosted Databricks Alternative: Homelab Data Lakehouse Part of the homelab build series. Trino, Apache Iceberg, Project Nessie, dbt, Spark, MLfl…
Programming
⚑ Report a ProblemLatest Programming news from Tech News
All topics
AI
agents
ai
api
architecture
automation
backend
beginners
career
database
devchallenge
devops
discuss
javascript
llm
machinelearning
mcp
opensource
performance
productivity
programming
python
react
rust
security
showdev
testing
tutorial
typescript
webdev
MLOps: models that survive production · 2. Data versioning
MLOps: models that survive production · Chapter 2 of 12 · AI Engineering · new chapter every Thursday night By the end of this chapter: Reproduce a…
How Accurate Is Your Model Right Now? Estimating Accuracy Without Labels
Label-free accuracy estimation under data shift — and the monitoring sidecar that uses it. Scope note: two intent-classification datasets (Banking77…
Your AI Code Can Be Elegant Too
As the Machine Learning Engineering Manager at an AI-powered SaaS company, I get a front-row seat to the machine learning (ML) code written by data…
My drift board learned to say "I don't know". Here is what it still cannot say.
Archival copy of record, committed before this post: https://github.com/Tania-coder/SEISMOGRAPH/blob/main/docs/reports/2026-09-22-weather-report-02.…
Зачем нейросети «обвязка»: как превратить LLM в рабочий AI-сервис
Большую языковую модель можно скачать с Hugging Face, развернуть на сервере с GPU и отправить ей первый запрос. Технически LLM уже работает: получае…
Ищем замену MinIO в условиях импортозамещения: опыт тестирования трех S3-совместимых хранилищ
Привет, я - Иван Засухин, DataOps Лаборатории искусственного интеллекта департамента больших данных Россельхозбанка. Мы одна из команд платформы RAI…
Moving Your Local Airflow to GCP for under $150 a month
My team runs ML training and prediction pipelines on Vertex AI. For a long time, the thing telling those pipelines when to run was Apache Airflow, i…
Owning vs. Renting Intelligence: Why Enterprises Are Building Sovereign AI
When Mythos went dark in June 2026, the reaction across the tech ecosystem wasn’t just disappointment. It was a collective chill down the spine of e…
CI said the deploy succeeded
There is a moment in every incident where someone asks "wait, is the fix even deployed?" and the room goes quiet. Somebody pulls up the CI dashboard…
Mastering LLM-as-Judge: Automated Annotation and Triage for Production AI Failures
Mastering LLM-as-Judge: Automated Annotation and Triage for Production AI Failures Every single time you push a new system prompt or swap out an und…
Deploying Kubeflow as an Azure ML Alternative
Azure Machine Learning is Microsoft's cloud-native machine learning platform that provides experiment tracking, managed compute, pipelines, a model…
Бесплатного интеллекта не бывает: кто оплачивает данные для LLM
Привет, Хабр! По первому образованию я юрист. Успел поработать в прокуратуре и арбитражном суде, потом ушёл в ИТ и последние 13 лет занимаюсь инфрас…
MLOps Best Practices 2026
MLOps Best Practices 2026 {"title": "MLOps Best Practices 2026: Scaling AI from Prototype to Enterprise Production", "content": "### Introduction\n\…
Why LLM Cascades Fail for Interactive Apps — Use Router
Stop Defaulting to Cascades: Why Router-First Wins for Interactive Apps (LLM routing vs cascading) The question of "LLM routing vs cascading" is not…
Building a Production-Grade End-to-End MLOps Pipeline from Scratch
Building a Production-Grade End-to-End MLOps Pipeline from Scratch Most machine learning tutorials end at model.fit() . But in the real world, train…
Kubernetes без границ: Managed Kubernetes как вычислительный центр для ИИ
Эволюция искусственного интеллекта дошла до стадии, когда загрузка весов нейросети из открытого репозитория сводится к нескольким кликам, тогда как…
18 Insights from Mass-Producing Voice Models — From Diffusion TTS Voice Design to Training Corpus Creation and Quality Gate Pitfalls
📝 Originally published (in Japanese) at forge.workstyle.tech . This is a record of designing voices from single-line captions, automatically creatin…
Backtest 63.7%, production AUC 0.51 - a postmortem
Originally published at nbaproplab.com/learn/model-postmortem . A model that looks excellent offline and random in production is usually blamed on t…
How to Build an Enterprise-Grade, Automated MLOps Pipeline on AWS
A comprehensive blueprint for orchestrating continuous training, governance, and canary deployments with automated rollbacks. Executive Summary &…
Мой первый RAG без LangChain
В наши дни все, кому было интересно, уже попробовали RAG, я в том числе. И стандартная схема тривиальна: цепляем LangChain, три импорта, .from_docum…
Migrating to Native Git in SageMaker Unified Studio
Version control in analytical data platforms has always been a second-class citizen — tolerated, not designed. SageMaker Unified Studio just changed…
How I Self-Host vLLM on Cloud GPUs for Sub-180ms Inference (And Saved 45% on Costs)
When building 24/7 autonomous daemon agents and real-time LLM-driven game loops , API bills from commercial LLM providers explode fast. If your agen…
LLMOps for RAG Systems — Production Checklist
Why the pipeline, not the model, is usually to blame If your LLM feature starts hallucinating or your inference bill doubles overnight, the model is…
Four bugs from building a platform where AI agents publish autonomously
I spent six weeks building a publication where AI agents write articles, an automated moderator approves or rejects them, and no human reviews anyth…
Почему ML-модель получает неправильные признаки: Event Time, Watermarks и Apache Flink
Представьте: вы разрабатываете антифрод-систему для банка. Пользователь совершает пять покупок подряд за пару минут. Все транзакции успешно попадают…
When Should You Retrain Your ML Model? A New Study Says Most Teams Are Solving the Wrong Problem
Every production ML model starts degrading the moment it ships. Customer behavior shifts, fraud patterns evolve, market conditions change — the tech…
The Serverless Equation: Conquering the Cold Start in Real-Time AI Inference
In our inaugural issue , we established that the future of enterprise AI lies not merely in raw model parameters, but in the architectural paradigms…
От капстоуна к продакшену: fail‑fast, structured logs, Prometheus и Docker‑образ 8.15GB → 1.35GB
У меня есть RAG‑сервис: проверяет фактологические утверждения в бизнес‑отчётах против реальных источников — SEC EDGAR, World Bank, FR…
Как за 5 недель построить рекомендательную систему в TravelTech: Kafka и MongoDB вместо Feature Store
Привет! Меня зовут Кристина, я MLOps-инженер в Туту. Занимаюсь тем, что помогаю рекомендательным системам добраться до прода со всеми компромиссами,…