Ai Network Architect

Inspiration

Our idea came from combining our different interests and experiences. Nyles previously interned in networking and systems engineering, while Samir is interested in backend development and previously interned in data engineering. We wanted to bring those areas together and explore how AI could make network design faster and easier without sacrificing reliability.

What it does

AI Network Architect turns a plain-English description into a complete network design, including VLANs, subnets, firewall rules, device configurations, and a parts list. Users can explore the generated topology, intentionally introduce configuration mistakes, simulate hardware failures, and see how traffic moves through the network.

How we built it

We use AI to turn unstructured user input into structured network requirements. From there, our backend builds a network model that powers the topology, device configurations, deterministic checks, failure simulation, reachability analysis, and parts list.

Challenges we ran into

One of our biggest challenges was deciding where AI should stop and traditional software should take over. We also had to carefully model network behavior so that every part of the application stayed consistent with the same generated design.

Accomplishments that we're proud of

We're proud that we were able to turn a single paragraph into an interactive network that users can explore, test, and experiment with. Seeing the topology, configurations, failure simulation, and reachability all work together in one system was one of the most rewarding parts of the project.

What we learned

We learned a lot about networking, backend architecture, data modeling, and integrating AI into a larger application. Most importantly, we learned that AI is most useful in this kind of system when it works alongside predictable, deterministic software rather than trying to handle everything itself.

What's next for AI Network Architect

We want to support more network types, vendors, and device configurations, while continuing to improve the simulator. In the future, we'd also like to connect generated configurations to network emulators or virtual devices for more realistic testing. We also see strong potential for AI Network Architect as both a learning tool for students and a design and testing tool for networking professionals.

Built With

• api
• claude
• css3
• fastapi
• html5
• javascript
• pydantic
• pytest
• python
• svg
• uvicorn