Inspiration

I’ve worked in restaurants. I’ve seen it firsthand — you ask an owner what customers are saying, and they shrug. “Check our Yelp,” they say.

But think bigger: what if you’re a senior manager overseeing 3–4 restaurants across the state? Maybe more. Are you really going to sit there and read through hundreds of scattered Google reviews? In an industry where “the customer is always right,” how are you supposed to actually listen to the customer?

That’s the real problem. Review platforms are built for consumers, not operators. They're full of noise — vague comments, duplicate complaints, generic 5-star ratings.

But buried in there? Actual gold. If you know how to dig it out.

We wanted to fix that.

So we asked ourselves: What if restaurants could get the kind of customer insight that a 5-star hotel chain gets — but without a six-figure data team? What if it was dead simple, dirt cheap, and delivered insights that actually help you run your business better?

That became Grubsy.

What We Learned

  • Restaurant owners don’t want another dashboard. They want straight-up answers. “What’s broken?” “What’s working?” “What should I fix tomorrow?”
  • If you offer a small discount, customers will happily leave a structured survey — and the insights from that are worth 100x more than a random Yelp comment.
  • Online reviews are a hidden goldmine, if you know how to mine them. NLP + opinion clustering is how we dig it up.

How We Built It

  • Google Reviews Scraper built with Node.js (yes, we handled rate-limiting and dynamic content)
  • FastAPI backend that processes and analyzes reviews using OpenAI and TensorFlow
  • Flutter Web frontend with Hive for local persistence (so refreshing the page doesn’t wipe insights)
  • Firebase Hosting for simplicity and lightning-fast load times

Challenges

  • Scraping Google reviews without getting banned? That’s a dance. We learned every step.
  • Turning raw AI output into insights that are actually helpful to a business owner — not just pretty word clouds
  • Designing an experience for real people: restaurant managers, not engineers

We built it. We shipped it.
Now we’re running pilots with real restaurants, collecting feedback, and improving fast.

The goal? Make customer feedback the most valuable asset a restaurant has — not a mess they ignore.

Built With

Share this project:

Updates