ElectrifyAI | Devpost

Inspiration

As the world rapidly advances toward powerful software-based AI assistants, we noticed a clear gap: hardware and electrical engineering workflows are still largely manual, fragmented, and inaccessible to intelligent assistance.
ElectrifyAI was inspired by the question: why shouldn’t circuit designers get the same AI leverage that software developers already enjoy?

What it does

ElectrifyAI is an AI copilot for electrical design that understands schematics, components, and user intent.
It helps engineers:

Interpret and reason about schematics
Suggest components and design fixes
Bridge natural language requests with real design actions

How we built it

We built ElectrifyAI using a modular agent-based architecture:

A frontend UI embedded into EDA tools
A backend reasoning layer that processes user intent, schematic snapshots, and component catalogs
Structured JSON-based input/output pipelines to ensure secure execution
Separation between reasoning, execution, and visual feedback to keep the system reliable and extensible

Challenges we ran into

Translating ambiguous human intent into hardware actions
Synchronizing multiple agents and ensuring downstream systems
Designing a schema that was flexible enough for AI reasoning but strict enough for execution
Working across software heavy tools as electrical engineering students

Accomplishments that we're proud of

Successfully getting an AI system to understand schematic context
Designing a clean, scalable I/O contract for AI-hardware interaction
Building a working end-to-end copilot pipeline
Proving that AI-assisted hardware design is not only possible, but practical

What we learned

Hardware AI assistants require far more structure and safety than software ones
Clear boundaries between reasoning and execution are critical
Good system design matters more than model size
Electrical engineering desperately needs better developer tools

————————————————————————————————————————- ——