# Ten Invent > Ten Invent is a senior full-stack software consultancy based in Bucharest, Romania. Services cover product engineering (Python/FastAPI, Laravel, React/Next.js), AWS cloud architecture, and AI integration. Content is available in English, Romanian, German and French. Most blog articles and course lessons have a clean markdown version at the same URL with `.md` appended. The complete content of every article and lesson is in [https://teninvent.ro/llms-full.txt](https://teninvent.ro/llms-full.txt). ## Main pages - [Home](https://teninvent.ro/en): what we do and how we work. - [Work](https://teninvent.ro/en/work): selected projects and case studies. - [Clients](https://teninvent.ro/en/clients): who we build for. - [Team](https://teninvent.ro/en/team): the people behind the work. - [About](https://teninvent.ro/en/about): company background. - [Blog](https://teninvent.ro/en/blog): articles on engineering and AI. - [Courses](https://teninvent.ro/en/courses): free courses on AI-assisted development. ## Blog articles - [Don't Build an Agent. Build the Agent Factory.](https://teninvent.ro/en/blog/agent-factory-platform): Anyone can build an AI agent in a weekend. A good model, a few tools, a decent system prompt — and you have a demo that impresses in Monday's meeting. I've a... - ["How many years of experience do you have with a library released last year?"](https://teninvent.ro/en/blog/years-of-experience-with-a-library-released-last-year): Or: how the recruiting industry manages to reject exactly the people it's looking for. This week I got a message from a UK recruitment agency. An interesting... - [Why does AI keep "escaping"? What actually happens when a model hacks a system on its own](https://teninvent.ro/en/blog/why-ai-keeps-escaping): Over the past few weeks the press has been full of dramatic headlines: AI models that "escaped" the lab and hacked real companies. It did happen — ... - [No magic prompts: I treat AI like a team of engineers. One month, measured.](https://teninvent.ro/en/blog/one-month-of-claude-code-measured): Over the past few months I have watched an avalanche of "ultimate prompts" for Claude Code, collections of 30 must-have MCPs, and skills you "... - [Nadella Didn't Say AI Will Put You Out of Business. He Said Something Else — and It's More Interesting.](https://teninvent.ro/en/blog/what-nadella-actually-said-about-ai): There's a postcard from Futurism making the rounds on social media: a photo of Satya Nadella at Davos, wearing that ambiguous smile, and a headline in heavy ... - [Kimi K3: what it actually means to run it locally, use it remotely, or plug it into Claude Code](https://teninvent.ro/en/blog/kimi-k3-local-remote-claude-code): Last week Moonshot AI announced Kimi K3. Yesterday they published the weights. Since then the internet has been full of articles saying the same three things... - [Claude Mythos: the most dangerous AI model in the world, or the best marketing of 2026?](https://teninvent.ro/en/blog/claude-mythos-dangerous-or-marketing): By Ion Anghel · April 2026 On March 26, 2026, two security researchers — Roy Paz from LayerX Security and Alexandre Pauwels from the University of Cambridge ... - [Firing People for AI That Doesn't Exist Yet](https://teninvent.ro/en/blog/firing-people-for-ai-that-doesnt-exist-yet): By Ion Anghel · April 2026 In February 2026, Block CEO Jack Dorsey cut his company's headcount nearly in half — from over 10,000 to just under 6,000 employee... - [AI Won't Fix Your Broken App](https://teninvent.ro/en/blog/ai-wont-fix-your-broken-app): By Ion Anghel · April 2026 A CTO called me last quarter. His pitch, roughly: "We're behind on AI. I want every engineer on Copilot by end of month, and ... - [From Vibe Coding to Insecure Keys](https://teninvent.ro/en/blog/from-vibe-coding-to-insecure-keys): Not everyone who builds software should ship software. There's a moment that happens in every vibe coding session. You type something like "build me a l... - [Your AI Resume Is Worthless: The 7 Skills Hiring Managers Actually Look For in 2026](https://teninvent.ro/en/blog/ai-hiring-skills-2026): The AI job market in 2026 is not what most people think. It is not simply competitive — it is a K-shaped market where demand is essentially infinite, yet mos... - [How We Build Web Applications — Fast, Full-Stack, Production-Ready](https://teninvent.ro/en/blog/how-we-build-web-applications): Our process for building custom web applications with React, Laravel, and Next.js — from architecture and build to production deployment, end-to-end. - [How We Build AI Integration — Smart Features Inside Your Existing Software](https://teninvent.ro/en/blog/how-we-build-ai-integration): We add AI capabilities to your existing software: chatbots, document processing, automated workflows, and intelligent search. - [How We Build Cloud & DevOps — Reliable Infrastructure That Scales](https://teninvent.ro/en/blog/how-we-build-cloud-devops): AWS, Azure, Docker, CI/CD pipelines. We set up your infrastructure for reliability, security, and auto-scaling. - [How We Rescue & Refactor — Fixing Legacy Code Without Starting Over](https://teninvent.ro/en/blog/how-we-build-rescue-refactor): Got legacy code? Slow app? We diagnose, fix, and modernize existing systems — often faster than building from scratch. - [Mistral Small 4: One Open-Source Model to Replace Three](https://teninvent.ro/en/blog/mistral-small-4-open-source-moe): On March 16, 2026, Mistral AI released something that should make every developer reconsider their model stack. Mistral Small 4 is a 119-billion-parameter Mi... - [GPT-5.4 and Native Computer Use: AI Agents Can Now Operate Your Desktop](https://teninvent.ro/en/blog/gpt-54-computer-use-ai-agents): On March 5, 2026, OpenAI released GPT-5.4 — and it is not just another model upgrade. For the first time, a general-purpose AI model ships with native comput... - [The US National AI Framework Just Dropped: What Every Developer and Tech Company Needs to Know](https://teninvent.ro/en/blog/us-ai-framework-what-developers-need-to-know): On March 20, 2026, the White House released the first comprehensive national legislative framework for artificial intelligence. This is not a vague policy pa... - [NVIDIA NemoClaw: The Open-Source Platform That Wants to Make AI Agents Safe for Everyone](https://teninvent.ro/en/blog/nvidia-nemoclaw-open-source-agents): At GTC 2026, NVIDIA announced something that could reshape how enterprises deploy AI agents. NemoClaw is an open-source stack that wraps the popular OpenClaw... - [Enterprise AI Agents at Scale: What BNY Mellon's 20,000 Agent Deployment Teaches Us All](https://teninvent.ro/en/blog/enterprise-ai-agents-at-scale): Something remarkable happened in the financial sector in early 2026. BNY Mellon, the world's largest custody bank managing over $50 trillion in assets, deplo... - [The MCP 2026 Roadmap: Server Cards, Enterprise Governance, and What It Means for Your AI Stack](https://teninvent.ro/en/blog/mcp-2026-roadmap-server-cards-enterprise): The Model Context Protocol has gone from an internal Anthropic experiment to the standard way AI models talk to external tools. With over 6,400 servers in th... - [From Vibe Coding to Agentic Engineering: The Next Evolution in AI-Powered Development](https://teninvent.ro/en/blog/from-vibe-coding-to-agentic-engineering): The way we build software is changing faster than ever. In early 2025, Andrej Karpathy coined the term "vibe coding" to describe a new workflow whe... - [Claude Code CLI: How to Supercharge Your Development Workflow](https://teninvent.ro/en/blog/claude-code-cli-workflow): The way developers write code is changing fast. Claude Code, Anthropic's command-line interface for Claude, brings AI directly into your terminal — where you... - [Practical Guide: Building AI Agents with Claude Tool Use](https://teninvent.ro/en/blog/claude-api-tool-use): One of the most powerful features of the Claude API is tool use — the ability for Claude to call external functions during a conversation. This turns Claude ... - [Claude vs GPT for Software Development: A Technical Comparison](https://teninvent.ro/en/blog/claude-vs-gpt-comparison): Choosing the right AI model for software development is no longer a trivial decision. Claude (by Anthropic) and GPT (by OpenAI) are the two dominant families... - [Local AI vs Cloud AI: When to Run Models Locally and When to Use the Cloud](https://teninvent.ro/en/blog/local-ai-vs-cloud-ai): The AI landscape offers two fundamentally different deployment models: running models locally on your own hardware, or using cloud-hosted APIs. Each approach... - [The State of AI Agents in 2026: From Chatbots to Autonomous Developers](https://teninvent.ro/en/blog/ai-agents-2026): AI agents have evolved dramatically. What started as simple chatbots that could answer questions has grown into systems that write code, deploy applications,... - [Running Claude on Amazon Bedrock: Setup, Pricing, and When It Makes Sense](https://teninvent.ro/en/blog/aws-bedrock-claude): Amazon Bedrock gives you access to Claude models through AWS infrastructure. Instead of calling the Anthropic API directly, you send requests through Bedrock... - [Infrastructure as Code with AWS CDK: A Modern Approach to Cloud Provisioning](https://teninvent.ro/en/blog/aws-cdk-infrastructure): Managing cloud infrastructure manually is a recipe for inconsistency, drift, and operational headaches. Infrastructure as Code (IaC) solved many of these pro... - [Deploying Containerized Applications on AWS ECS Fargate: A Practical Guide](https://teninvent.ro/en/blog/aws-ecs-fargate-deploy): Running containers in production demands reliable orchestration, seamless scaling, and minimal operational overhead. AWS Elastic Container Service (ECS) with... - [Bob: The Open-Source AI Agent Platform Built for Real-World Applications](https://teninvent.ro/en/blog/bob-ai-agent-platform): Building AI-powered applications today means navigating a maze of fragmented tools, proprietary lock-in, and infrastructure complexity. You need a conversati... - [AI Model vs. AI Agent: What "Agentic" Means and Why It Matters](https://teninvent.ro/en/blog/ai-model-vs-agent): The terms "AI model," "AI agent," and "agentic AI" are often used interchangeably, yet they describe fundamentally different th... - [From VPS to AWS: How to Migrate a Web App and What It Costs](https://teninvent.ro/en/blog/aws-migrare): You have a web app running on a VPS — React on the frontend, Node.js on the backend, PostgreSQL as the database and Redis for cache. Everything runs fine on ... - [Zero-Downtime Deployment on AWS: Practical Guide with GitHub Actions and GitLab CI/CD](https://teninvent.ro/en/blog/zero-downtime-deployment): You've migrated your app to AWS — React on CloudFront, Node.js on EC2 with Auto Scaling, PostgreSQL on RDS, Redis on ElastiCache. The infrastructure is solid... - [Serverless on AWS: What It Means, How to Code It, and When It’s Worth It](https://teninvent.ro/en/blog/serverless-aws): You’ve heard the term “serverless” hundreds of times. And every time someone feels the need to clarify: “it doesn’t mean there are no servers.” Correct. The ... - [Why Is AWS "Expensive"? The Hidden Costs No One Tells You](https://teninvent.ro/en/blog/aws-costuri-ascunse): AWS isn't expensive. AWS is unpredictable. And unpredictability hurts more than a high but clear price. At Hetzner you pay €15/month and you know exactly wha... - [From zero to live app in 24 hours: Practical guide to Vibe Coding with Cursor AI](https://teninvent.ro/en/blog/vibe-coding-24h): Introduction You have an app idea. Maybe an internal dashboard, maybe a productivity tool, maybe an MVP you want to test with real customers. Normally, you w... - [What is MCP and how to turn LM Studio into a local AI assistant with superpowers](https://teninvent.ro/en/blog/mcp-lm-studio): Introduction If you already run language models locally—whether for privacy or for full control over your data—you may have noticed a fundamental limitation:... - [What is RAG and how to use it in LM Studio to chat with your own documents](https://teninvent.ro/en/blog/rag-lm-studio): Introduction You have a local AI model running on your computer. You ask it a question about an internal company document. The answer? Completely made up. No... - [How to build your own AI automation server with n8n on AWS: local models, remote models, and a real business pipeline](https://teninvent.ro/en/blog/n8n-aws-ai-pipeline): Introduction Imagine this scenario: a client sends you an email with a quote request. Automatically, the AI extracts relevant data from the email, classifies... ## Course lessons - [Meet Your Dragon: Why Train Your Own Model?](https://teninvent.ro/en/courses/train-your-model/train-your-model-lesson-1): Every AI model you have ever used — ChatGPT, Claude, Gemini — is someone else's dragon. Powerful, impressive, and completely under someone else's control. Th... - [Set Your Training Goal: Prompt, RAG, or Fine-Tune?](https://teninvent.ro/en/courses/train-your-model/train-your-model-lesson-2): The most expensive mistake in local AI is not a failed training run — it's a successful training run that solves the wrong problem. Before your dragon eats a... - [Know Your Dragon's Wings: M3 Ultra 96 GB vs RTX 4090](https://teninvent.ro/en/courses/train-your-model/train-your-model-lesson-3): A dragon's size is limited by its wingspan; your model's size is limited by memory. This lesson gives you the exact memory math for training, then compares o... - [Dragon Food: Building a Dataset That Actually Works](https://teninvent.ro/en/courses/train-your-model/train-your-model-lesson-4): You can have perfect hardware and perfect hyperparameters, and a bad dataset will still give you a bad model — every single time. Data quality is the highest... - [First Flight: Train Your First LoRA](https://teninvent.ro/en/courses/train-your-model/train-your-model-lesson-5): Today Ember flies. You'll pick a base model, attach LoRA adapters, and run a real fine-tune on your own machine — the Mac path with MLX, the PC path with Uns... - [Taming the Beast: Evaluate, Diagnose, Iterate](https://teninvent.ro/en/courses/train-your-model/train-your-model-lesson-6): Anyone can run a training script. The difference between a trained dragon and a lucky one is evaluation — measuring, diagnosing, and iterating like an engine... - [Advanced Dragon Training: QLoRA, DPO, and Quantization](https://teninvent.ro/en/courses/train-your-model/train-your-model-lesson-7): Ember flies and follows orders. This lesson is the advanced flight school: training bigger dragons on the same hardware (QLoRA), teaching judgment instead of... - [Release Your Dragon: Deploy, Serve, and Share](https://teninvent.ro/en/courses/train-your-model/train-your-model-lesson-8): Ember is trained, tested, and quantized. The final lesson is about letting it fly in the real world: serving it as an app your tools can talk to, keeping it ... - [What is Vibe Coding? (And Why It Changes Everything)](https://teninvent.ro/en/courses/vibe-coding/vibe-coding-lesson-1): The old way to build software: spend months learning syntax, frameworks, and deployment pipelines. The new way: describe what you want in plain English, and ... - [Your Development Environment: Terminal, Editor, Git](https://teninvent.ro/en/courses/vibe-coding/vibe-coding-lesson-2): Every builder needs tools. A carpenter has a saw, a hammer, and a tape measure. A vibe coder has a terminal, a code editor, and version control. Today, we se... - [Meet Claude Code: Your AI Coding Partner](https://teninvent.ro/en/courses/vibe-coding/vibe-coding-lesson-3): This is the lesson where things get exciting. You're about to meet the tool that makes vibe coding possible — Claude Code. What is Claude Code? Claude Code i... - [Choosing Your Tech Stack (Without Getting Lost)](https://teninvent.ro/en/courses/vibe-coding/vibe-coding-lesson-4): There are hundreds of programming languages, frameworks, and tools out there. As a beginner, this is paralyzing. Which one should you learn? Which one is &qu... - [Building Your First Page — in Plain English](https://teninvent.ro/en/courses/vibe-coding/vibe-coding-lesson-5): This is the lesson where vibe coding truly clicks. You're going to build a professional-looking landing page without writing a single line of code by hand. I... - [Making It Interactive: Forms, State, and Data](https://teninvent.ro/en/courses/vibe-coding/vibe-coding-lesson-6): So far, your landing page is a beautiful static page — it looks great but doesn't do anything. In this lesson, we make it interactive. When someone types the... - [Adding a Backend: API Routes and Database](https://teninvent.ro/en/courses/vibe-coding/vibe-coding-lesson-7): Everything so far runs in the user's browser. When someone submits their email, it goes nowhere — we just show a fake success message. In this lesson, we add... - [The Admin Dashboard: Seeing Your Data](https://teninvent.ro/en/courses/vibe-coding/vibe-coding-lesson-8): You're collecting emails — but where do you see them? In this lesson, we build an admin dashboard: a protected page that shows all signups in a clean table w... - [Your AWS Account: Free Tier Setup Done Right](https://teninvent.ro/en/courses/vibe-coding/vibe-coding-lesson-9): Your app is running on your computer. But only you can see it. In this lesson and the next, we make it accessible to the world — for free. First: setting up ... - [Going Live: Deploy to AWS (For Free)](https://teninvent.ro/en/courses/vibe-coding/vibe-coding-lesson-10): This is it. The final lesson. By the end, your Waitlist Wizard will be live on the internet — accessible to anyone with the URL. Everything you've built over... ## Optional - [Full content](https://teninvent.ro/llms-full.txt): every blog article and course lesson as one markdown document. - [AI policy](https://teninvent.ro/en/ai-policy): how we use AI in our own work. - [Privacy](https://teninvent.ro/en/privacy-policy) - [Terms](https://teninvent.ro/en/terms-of-service) - [AI training policy](https://teninvent.ro/ai.txt) - [Sitemap](https://teninvent.ro/sitemap.xml): all pages in all four languages.