Machina

Inspiration

It all started with frustration. I was researching suppliers for a hardware project and realized how outdated and opaque the whole manufacturing ecosystem still is. Every site I visited IndiaMART, TradeIndia, even global ones felt like stepping into a time capsule from 2007. Prices were inconsistent, supplier info was scattered or missing, and trying to understand the actual cost of a finished product vs. its components was a guessing game. There was no transparency, no intelligence, and zero trust built into the process. I kept thinking: "Why isn’t there a better way to source parts, benchmark costs, and evaluate supplier quality like how we do with modern SaaS tools?" That thought wouldn’t leave my head and that’s where Machina was born.

What it does

Machina is a voice-powered manufacturing Assistance hub. It brings together a global directory of parts, products, and suppliers with tools that calculate meaningful indices like the “Idiot Index” (how overpriced a finished product is compared to its parts) and the “Quality Rate Index” (based on supplier performance). The goal is simple: remove the guesswork from procurement, help manufacturers vet suppliers better, and turn sourcing into a strategic edge not a bottleneck. The best part? You can talk to Machina like a teammate. It responds in natural language and helps you find answers instantly, from BOM comparisons to suggesting alternate suppliers.

How we built it

I combined a full-stack web app using Next.js and React for the frontend, Supabase for the database and auth, and ElevenLabs for the voice interface. I paid attention to every detail from product breakdowns and supplier data to how the voicebot feels when you talk to it. The UI design is heavily inspired by Apple’s clean, precise aesthetic. I wrote serverless functions to calculate pricing indices and made sure users could get real-time insights without needing a spreadsheet or sourcing consultant. It’s all built with transparency and clarity at the core.

Challenges we ran into

Sourcing reliable component-level data was a pain. I had to think hard about how to normalize pricing info, deal with incomplete BOMs, and calculate indices in a way that made sense across different industries. Getting the voice assistant to give smart, contextual answers (instead of just generic replies) also took a lot of iteration. And of course, making a UX that felt modern while working with manufacturing data (which is usually messy and dry) was a big design challenge.

Accomplishments that we're proud of

Turning something that used to feel like a dead-end Google Sheet into a dynamic, intelligent platform you can talk to feels huge. The Idiot Index and QRI aren’t just gimmicks they’re a way for people in the manufacturing world to make smarter, faster, and more transparent decisions. I’m proud of building something that makes the backend of innovation (manufacturing) feel as frictionless and modern as the frontend (apps, AI, etc.).

What we learned

Manufacturing isn’t broken it’s just underserved. The people in this industry are resourceful and sharp, but they’re stuck using outdated tools and workflows. If you build something genuinely useful, with empathy and clarity, they’re more than ready to use it. I also learned how important it is to bridge UI/UX with domain complexity. It’s not enough to show data you have to make it feel usable and trustworthy.

What's next for Machina

Next, I want to go deeper into data partnerships so we can bring even more verified supplier information and component pricing to the platform. I’d also like to explore on-the-fly BOM analysis, better quote negotiation tools, and more languages/localization. But the core idea remains the same: give manufacturers and suppliers tools that don’t just support their work but actually think with them.

Built With

• authentication
• backend
• cloud
• elevenlabs
• framer
• languages:-typescript-javascript-html-css-frontend-frameworks-&-libraries:-react-react-router-dom-(for-client-side-routing)-radix-ui-(for-unstyled
• lucide
• postcss
• postgresql
• services:
• supabase
• typescript
• vite