Latest Tutorials

In this series of articles we’re going to be talking about how to use LLMs to generate synthetic data for QA testing, starting with the basics of test data, then moving on to generation methods, and finally looking at examples for generating test data for the purpose of validating LLM products. But let’s start at the beginning - in this article we’re going to talk about how to use synthetic test data more generally, what makes good or bad test data, and we’ll also look at some traditional QA methodologies and how test data can inform them. Synthetic data refers to any machine-generated data that can be used to execute test cases or mock a production environment scenario. This includes data produced by LLMs, procedural data, and human curated or created data generated outside of production. Of course, production data is incredibly valuable for testing, and when it’s possible to use it, it should be used - but often this is not possible, legal or scalable. Generating production data can also be an expensive process for a new feature or product since you need to hire beta testers. Synthetic data also has some other advantages other than cost.

Imagine having the power to gather data from any website—product prices, news articles, or social media posts—with just a few lines of code. Maybe you’ve tried building web scrapers before — only to hit endless walls with anti-bot protections, slow speeds, or messy data formats. In today’s AI-driven world, simply grabbing raw HTML isn’t enough. We need clean, structured, and meaningful data that’s ready for machine learning models, data pipelines, and AI agents to understand. That’s where Crawl4AI steps in. In this beginner-friendly guide, we’ll walk you through the basics of Crawl4AI, an open-source Python tool designed to scrape and extract web data effortlessly. You’ll learn what it is, why it’s a game-changer, and how to set up your first web crawler step by step.

Ever feel like you’re drowning in repetitive tasks? Sorting emails, posting updates, or organizing files can eat up hours of your day. But what if you could automate all of that — and have an AI assistant that actually understands what you need? That’s where n8n comes in. n8n is an open-source, low-code automation platform that not only helps you connect your favorite apps and tools but also gives you access to powerful AI features. Whether it's managing your to-do list or answering complex queries with a single prompt, n8n lets you build smart, automated workflows with ease.

Learn how to create a custom Typeform-style survey with Replit Agent, an advanced AI-coding agent, and Notion, a popular productivity note-taking application. This step-by-step guide teaches you how to go from an initial idea to an animated, conversational survey form in minutes, regardless of skill level.

Welcome to the second article in our Model Context Protocol (MCP) series! In the first article , we covered all the basics—what MCP is, how it works, and its key components. Not only that, we even built our first simple MCP server to put that knowledge into practice. If you’re new to this topic, I highly suggest you check out that first article before continuing here. It’ll give you a clear understanding of what’s going on and make this journey a lot smoother. Now, it’s time to step things up. In this article, we’re going to take what we’ve learned and build something more advanced – a custom MCP server from scratch. This is where things get interesting, because we’ll see just how flexible and powerful MCP can be with the new tooling we’re going to explore, and how we can shape MCP to fit our own needs. Let’s get into it! First, let’s quickly mention the tech stack that we’re going to work with. We’re going to use the same one from the previous article: TypeScript, Node.js and MacOS. If you're using a different tech stack, no worries, the key ideas will be the same. As an additional reference, you can also refer to documentation which includes basics for Python and Java too.