NetConfig AI

500,000+ network engineers worldwide write Cisco IOS commands manually every day. Complex CLI syntax is hard to memorize, a single typo can bring down an entire network, and junior engineers spend weeks learning commands with no natural language interface available. NetConfig AI solves this. Users type what they want in plain English — for example, "Configure OSPF with network 192.168.1.0/24 in area 0" — and the AI instantly generates correct, production-ready Cisco IOS commands. The system remembers conversation context across turns, so users can iteratively build complex configurations: adding new OSPF areas, layering in MD5 authentication, then combining everything into one final config block. NetConfig AI also analyzes uploaded network diagrams, comparing multiple topologies and generating commands based on visual content. Once commands are generated, users can deploy them directly to a real router via SSH, or test safely first using a GNS3 virtual router simulator. Technologies used: The frontend is built with React, Vite, and TailwindCSS. The backend runs on Flask (Python), using Netmiko for SSH-based router deployment. AI inference is powered by Fireworks AI, running the Kimi K2.7 model on AMD GPU infrastructure, for both text-based command generation and vision-based diagram analysis. Firebase handles authentication, session management, and chat history storage, while Cloudinary manages network diagram uploads. The entire application is containerized with Docker. The result: what used to take a network engineer 30 minutes of manual CLI work now takes under 10 seconds.