



Source Code: https://www.theblockchaincoders.com/sourceCode/i-built-an-ai-agent-desktop-app-that-controls-the-browser-voice-scraping-and-multi-llm-support-full-tutorial Blockchain Course: https://www.theblockchaincoders.com/pro-nft-marketplace Private Blockchain Course: https://www.theblockchaincoders.com/build-private-blockchain-course All Project Code: https://www.theblockchaincoders.com/SourceCode Donate Please: https://linktr.ee/daulathussain 1 - 1 Consultancy: https://www.theblockchaincoders.com/consultancy Pro Blockchain Courses: https://www.theblockchaincoders.com/ Public Discord: https://discord.gg/Gah6YGuBFS I Built an AI Agent Desktop App That Controls the Browser — Voice, Scraping & Multi-LLM Support - Full Tutorial 🤖 Build a production-ready AI Desktop Agent from scratch using Electron, React, TypeScript, and Playwright! In this tutorial, I'll show you how to build a powerful desktop application that lets AI models autonomously control a real browser — navigating websites, clicking buttons, filling forms, scraping data, and completing complex tasks — all through simple natural language instructions. 📌 Timestamps 00:00:00 ➤ Introduction 00:00:21 ➤ Overview 00:07:55 ➤ Starter File 00:08:44 ➤ Final Source Code 00:14:21 ➤ Installation 00:15:39 ➤ Testing Live 🚀 WHAT WE BUILD - ✅ Full Electron desktop app with embedded live browser view - ✅ AI-powered browser automation using Playwright - ✅ Support for 5 LLM providers (OpenAI, Claude, Gemini, Ollama, LM Studio) - ✅ Voice commands using OpenAI Whisper - ✅ Voice narration — the agent speaks every step it takes - ✅ Web scraping with one click + Excel file export - ✅ Secure API key storage using system keychain - ✅ Human-like delays & mouse movements to avoid bot detection - ✅ Auto login detection & manual login pause/resume - ✅ CAPTCHA detection and pause - ✅ Real-time logs, step tracker, and plan viewer - ✅ Custom app icon & production build for Mac & Windows 🛠️ TECH STACK - • Electron 33 — Desktop app framework - • React 18 + Vite — UI with lightning-fast HMR - • TypeScript 5 — Type-safe throughout - • Playwright 1.48 — Real browser automation - • OpenAI SDK — GPT-4o + Whisper voice - • Anthropic SDK — Claude Opus / Sonnet / Haiku - • Google GenAI SDK — Gemini 1.5 / 2.0 - • Ollama / LM Studio — 100% local, offline AI - • SheetJS (xlsx) — Excel export from scraped data - • Keytar — OS-level encrypted API key storage - • Web Speech API — Text-to-speech agent narration 📌 KEY FEATURES EXPLAINED 🧠 Multi-LLM Support Switch between OpenAI, Claude, Gemini, Ollama and LM Studio with a single dropdown. Use cloud models or run 100% offline. 🌐 Embedded Browser A real Chromium browser is embedded inside the app window. Watch the AI navigate in real time, right inside your desktop app. 🗣️ Voice Commands + Voice Narration Speak your task instruction into the mic — Whisper transcribes it. Enable voice mode and the agent narrates every step it takes out loud. 📊 Web Scraping + Excel Export One-click scraping of any page — tables, lists, links, and text are extracted and saved to a formatted Excel file instantly. 🔐 Secure & Smart API keys are stored in your OS keychain — never in plain text. The agent detects login pages and pauses for you to log in manually. Save NFT Marketplace PlayList: https://youtube.com/playlist?list=PLWUCKsxdKl0olgEF4OxXVk2B-jwpGqL5d API PlayList: https://youtube.com/playlist?list=PLWUCKsxdKl0oAFAVuRZxQSYC07UTcl_v_ Solidity PlayList: https://youtube.com/playlist?list=PLWUCKsxdKl0oksYr6IG_wRsaSUySQC0ck Complete JavaScript Course: https://youtube.com/playlist?list=PLWUCKsxdKl0qROhA0XO4_ek9bIwZ4j4Xr HTML Course Code: https://www.daulathussain.com/complete-html-course-daulat-hussain/ =================== HOSTING ++++++++++++++++++++ Best Hosting: https://clients.domainracer.com/aff.php?aff=28826 Follow Me: Instagram: https://www.instagram.com/daulathussain92/ Facebook: https://www.facebook.com/daulat.hussain.18 Twitter: https://x.com/TheBCoders Pinterest: https://in.pinterest.com/daulathussainhealthfitness/ Linkedin: https://www.linkedin.com/in/daulat-hussain/ Quora: https://www.quora.com/q/schahkxkdudpgjvh Facebook Group: https://www.facebook.com/groups/59011 Facebook Page: https://www.facebook.com/yourdhfitness Subscribe to My Channel: https://www.youtube.com/channel/UCz6_...

Raise 0.001 SOL per SPL Token 🚀 | Launch Your Solana Crypto DApp 2026
17 views

I’m a Good Vibe Coder… So Why Did I Get Rejected in 2026? 🤯
67 views

Oracle Layoffs Shock 😳 | Tech Jobs Are NOT Government Jobs!
557 views

Blockchain Security Researcher & Auditor Course 2026 🔐 Ethical Hacking for Web3
140 views

I’m a Developer… Come Waste My Time 😤 (Reality of Coding Life)
1.4K views

You Are Being Fooled as a Developer (Stop Wasting Your Time!)
3.0K views