Habr AI
AI news source. Articles are auto-selected and adapted by Hamidun News editors.
Latest publications
The best code didn't win: what the Dev-to-Dev hackathon final says about agentive engineering
At the Dev-to-Dev hackathon final, a project with flawless code (65 tests, full type coverage, security) took second place. The winning solution was the one that best captured the essence of agentive engineering. This po
Process Mining: the tool that saves corporate AI budgets
Most corporate AI projects eat up millions but deliver no results — Process Mining helps make sense of processes and invest money wisely.
Yandex automated Chromium updates with an LLM agent
Every four weeks, when Chromium is updated, Yandex Browser developers have to resolve thousands of code conflicts. An LLM agent now helps automate the routine and saves several person-months of work.
AI video requires more manual work than the tools promise
A Habr author tried making videos with AI and found that the tools' polished promises do not match reality.
Where AI Really Helps Sales: 2026 Use Cases Instead of Disappointment
AI adoption in sales often disappoints because it starts with ChatGPT rather than strategy. We break down where AI saves hours and where it is better kept away from customer communication.
YOLO and Redis: how to cache video detection results
YOLO video processing recalculates frames from scratch after a network failure. Redis caching solves this: each frame is processed once, and the results are restored after a restart.
Cursor in enterprise development: an AI editor for rapid prototyping
How the team used Cursor to develop the Planning module, reducing tab switching and staying focused on code instead of routine tasks.
Deepseek-agent: how to write your own Claude Code in 2,000 lines of code
A developer created a fully functional Claude Code equivalent for DeepSeek with support for tool use, permissions, and memory. Just 2,000 lines of Node.js and four dependencies rather than frameworks.
How to save 70% of Claude tokens with targeted code access
A developer built a CLI tool in PHP + SQLite that creates a project graph and gives Claude targeted access to the required code instead of reading entire files.
How 15 AI models approach finding the optimal XML parser for iOS: benchmark results
A developer gave 15 popular AI models the same task of finding a fast XML parser for iOS and compared their results with his own three-hour manual benchmark — the outcome was unexpected.
Self-recovery metric: researchers created an ASI criterion for evaluating AI
Researchers proposed the ASI criterion, a new approach to evaluating AI models based on their ability to self-recover and develop from an informational 'seed'.
AI agent security in production: a practical guide to Red Teaming
AI agents with access to corporate data and tools require a special approach to security testing. Doubletapp shared a practical guide to Red Teaming: how to find vulnerabilities before deployment to production and reduce
Spring AI made it easier to integrate Java applications with Claude and other LLM providers
Spring AI provides ready-made techniques for integrating LLM providers into Java, including working with Claude from Anthropic. The article covers practical development examples and when this makes sense.
Claude AI resolves arbitration disputes in 12 minutes instead of weeks
An author created a Telegram bot on Claude Sonnet that analyzes court cases in 12 minutes. In three out of four tests, its decision matched the verdicts of human arbitrators.
Protecting LLMs from semantic-level attacks: why a traditional firewall is ineffective
Neural networks face new types of attacks that traditional firewalls cannot stop. AI/LLM Firewall operates at the semantic level, blocking 80+ techniques from the MITRE ATLAS framework.
RAGAS and RAG metrics: how to stop guessing and start measuring quality
The third part of our quality engineering series: how to automate the evaluation of RAG systems with RAGAS and monitor completeness, precision, and faithfulness of outputs instead of relying on manual review.
AI agent in one day: a local prototype without the cloud or developers
A practical guide for companies that want to quickly deploy an AI agent without the cloud, large budgets, or a team of developers: Ollama, n8n, and prompt engineering in one working day.
Rosenblatt's Perceptron: How the History of AI Began
The story of the first working neural network, the perceptron, created by Frank Rosenblatt in 1958, and its role in the beginning of the modern AI era.
How DocAI adapts medical education for each student
DocAI is a Russian AI-based project for medical education that automatically adapts content to each student's level and provides a personalized learning path.
AI image generators aren't creative: why and what to do
Illustrations for a text can be generated in a couple of minutes. But why does the result often look generic? Because neural networks choose statistically likely solutions, not creative ones.
How Creatures (1996) created the first true computer creatures
Steve Grand refused to write AI. Instead, he wrote genetics, emotion simulation, and neurons. The result became a cultural phenomenon.
Can LLMs find flaky tests from code? Study says no
A study found that LLMs are poor at identifying unreliable tests, even when dataset metrics suggest otherwise.
Computer vision platform that requires no GPUs or ML specialists
Developers unveiled a platform for training computer vision models on small datasets — a dozen images is enough to achieve 86% accuracy without GPUs or ML expertise.
Axera AX650N: Edge AI chip for local robotics instead of Jetson
Axera has unveiled the AX650N Edge AI chip for local neural network execution in robotics. An alternative to the costly NVIDIA Jetson, it runs YOLO, LLM and VLM directly on the device without the cloud.
LLMs write Rust with seven categories of hidden errors the compiler doesn’t catch
After six months of using Claude, GPT, and Cursor for Rust development, seven categories of errors were identified that the compiler misses but that compromise code safety or architecture.
Managing AI agents is now easier: the best open-source interfaces
User-friendly open-source tools with graphical interfaces have emerged for managing a fleet of AI agents. They help build task pipelines and distribute work among agents without programming.
How BERTopic with a local LLM helps Rostelecom analyze large text collections
An NLP engineer at Rostelecom shows a pipeline for clustering large volumes of text and generating interpretable topic names using BERTopic and a local LLM.
Several LLMs in one Telegram chat: how to combine Groq and Google AI
A developer created a Telegram bot where several LLM models work in a single chat and pretend to be corporate employees. Users switch between them on demand, using only free APIs.
Harness around an LLM delivers multi-fold gains: what changed after a year with Claude Code
An engineer who spent a year with Claude Code found that the harness around the model — system prompt, tools, context and skills — delivers a much larger quality gain than upgrading the model itself.
How Alpina Digital shielded AI from Russia's data law
88% of companies use AI, but only 1% have reached maturity. Alpina Digital found a solution: an architecture where requests go to the cloud while processing stays local. The result: 4.5x faster and $700,000 saved.