Vidya AI Is Here- Your Assisted AI To Learn AI From Basics

Summary

Krish Naik announces the launch of Vidya AI, a new assisted learning platform developed with his team over three to four months. The tool is designed to support students learning artificial intelligence and is hosted for free on the krishnaik.in website under the AI Tutorials section. To use the interactive AI assistant features, users must provide their own Google Gemini API key, which Naik notes can be obtained using initial cloud credits between 200 and 300 dollars on new accounts. Before diving into the platform walkthrough, Naik briefly mentions the AI Forward Deployed Engineer Bootcamp starting its first class.

Navigating to the AI Tutorials tab, Naik displays the structured course modules available on the platform, including statistics, machine learning, natural language processing, and deep learning, with additional courses on generative AI, agentic AI, and retrieval-augmented generation planned for future release. He opens the machine learning tutorial to show how lessons combine embedded YouTube videos, diagrams, and written theory covering algorithms like gradient descent and polynomial regression. The platform also catalogs tutorials for essential AI frameworks, including LangChain, LangGraph, Deep Agents, Model Context Protocol, DeepEval, RAGAS, NeMo Guardrails, Claude Code, LangMem, and Mem0. Naik emphasizes that framework tutorials will be updated at least once a month to match new documentation releases.

The video then covers the platform's role-based roadmaps, specifically highlighting tracks for Agentic AI Engineers and AI Forward Deployed Engineers. The roadmaps break learning into weekly stages such as Python for production, LLM engineering, and deployment. Naik opens a Python lesson to demonstrate the integrated in-browser coding lab, where users can write, edit, and run Python code alongside lesson text, viewing the terminal output directly in the interface.

Naik transitions to the interview preparation section, which features categorized question banks covering topics like RAG, LLMOps, AI system design, and evaluation guardrails. Each set contains 50 questions with 30-second spoken answer prompts. To showcase the flagship feature, Naik activates Vidya AI on a RAG interview question page. He initiates a live voice conversation with the tutor, demonstrating how Vidya AI absorbs the exact text and diagram context of the active page. When Vidya AI quizzes Naik on the risks of relying solely on pretrained models, Naik answers by voice, and the assistant confirms the response, explains vector embeddings, and prompts him with the next architectural step. Naik concludes by explaining that Vidya AI leverages over two thousand of his uploaded YouTube videos as instructional context to help learners articulate technical knowledge.