-
-
Dashboard showing AI vision, circuit validation, simulation, and PDF report generation.
-
Upload screen for circuit images or PDFs up to 20MB.
-
File selected and ready to start parsing.
-
AI recognition in progress extracting components and pin nets.
-
Processing circuit graph and checking pin net consistency
-
Review dashboard with confidence scores and manual verification alerts.
-
Table of extracted components with confidence scores per metric.
-
Verification summary and downloadable component manifest preview.
-
Plain text component inventory list with confidence ratings.
-
PDF report generated with validation findings and viewer.
-
Final statistics dashboard with component counts and circuit diagram.
Picture this. It is a Friday evening, and you are hunched over a sprawling electrical schematic. You are squinting at cryptic component labels, manually typing specs into a spreadsheet, and dreading the mountain of work ahead. Your weekend is officially over before it begins. That exact frustration sparked our journey.
Our team lives for challenges. We actively hunt for opportunities to stretch our skills and grow. So when we stumbled upon the Devpost AI Cloud API Hackathon, we knew we had to jump in. This was our arena. This was the perfect test to transform a gnarly real world headache into something elegantly simple.
Fueled by excitement and way too much caffeine, we started brainstorming. We hit dead end after dead end. Heated debates. Whiteboards covered in crossed out ideas. Finally, we narrowed it down to four solid concepts, all merging artificial intelligence with electrical engineering. But then we hit a brutal wall. Our research showed that every single one of those ideas already existed. It was deflating.
But then we looked closer at the how. The existing tools out there are basically dinosaurs. They demand heavy high end RAM, robust processors, and clunky local installations. They lock out anyone with a standard laptop. That is when our vision crystallized. We wouldn't build just another parser. We would build an online sanctuary. A website, not a software install. One click, not a lengthy setup. Any laptop, anywhere. That is our killer advantage and our path to outperforming the big guys.
We did not just dream; we built. We made serious progress crafting a flawless five step workflow. First, the user uploads their messy schematic. Second, our AI recognition engine locks onto every hidden component buried in those traces. Third, we give you control with a review step to double check low confidence picks because trust is earned, not given. Fourth, we run a full SPICE simulation to verify the circuit actually works. Fifth and finally, we harness the powerful Nutrient DWS API to wrap it all up into a pristine, professional engineering PDF report.
We walked into this hackathon with a blank slate and walked out with a fully functional, cloud powered solution. We tackled a universal engineering headache that haunts students, amateurs, and seasoned pros alike. And with our scalable web architecture crushing the limitations of legacy local software, we are not just building a tool. We are building a real business ready to scale. This is not the finish line. This is just the beginning of something massive.
Built With
- altium
- amazon-web-services
- dws
- fastapi
- gemini
- llms
- nemotron
- next.js
- opencv
- openrouter
- pillow
- python
- react
- reportlab
- tailwindcss
- typescript
Log in or sign up for Devpost to join the conversation.