We look past the easy demos and headlines to understand what's happening underneath. Our focus is on the practical mechanics: fine-tuning models, rigorous evaluation, and the detailed work of prompt engineering.

This is a space for hands-on experimentation. We build, test, and sometimes break things to learn how AI behaves in the real world, especially when tackling harder problems.

Think of this as a garage where a bunch of friends experiment with AI tools, and document the journey while others are using them for making faceless YouTube videos and TikToks.

What you'll find here

• Detailed guides on fine-tuning techniques and dataset preparation.
• Deep dives into writing effective system prompts and evaluating model performance.
• Honest explorations of the capabilities and limitations of diverse AI models.
• Practical code examples and insights from hands-on experiments.

What you won’t find here

• Quick hacks or surface‑level summaries
• Tips on how to use AI to make quick money
• Clickbait "end of the world due to AI" articles and threads
• An AI course or community

We're driven by curiosity and the desire to build real understanding. We share substantive insights for fellow builders, researchers, and anyone who wants to look beyond the AI surface.

About the founder

Karto Labs is founded and run by Karthik Pasupathy. He explores the practical application and deeper implications of AI systems, sharing insights from his hands-on work in India.