I built an AI upselling system for restaurants after visiting 47 cafes in Bangalore — just launched on Product Hunt today
Hey IndieHackers,
I want to share the story behind Orlena and would
love honest feedback from this community.
THE PROBLEM I KEPT SEEING
Over the last year I visited 47 cafes and restaurants
across Bangalore Koramangala, Indiranagar, HSR Layout,
Jayanagar.
Every single one had a QR menu.
Not a single one had a QR menu that actually sold
anything.
Here is what I kept observing:
Servers were too busy to upsell consistently. They
missed upsell opportunities at over 70% of tables
not because they were bad at their jobs, but because
they were managing 8 tables at once.
High-margin items like desserts, sides, and premium
drinks sat buried at the bottom of the menu. Customers
defaulted to what they already knew.
And the QR menu, the technology the whole industry
moved to after COVID, was doing absolutely nothing
to fix this. It was just a digital PDF.
For a cafe serving 80 tables a day, this meant losing
Rs 50,000 to Rs 2,00,000 every single month in missed
revenue. From the same customers they already had.
THE SOLUTION
I built Orlena.
Orlena is an AI-powered QR menu upselling system.
When a customer scans the QR code and browses the
menu, Orlena's AI automatically recommends high-margin
add-ons at the exact moment they are ordering.
Customer orders a cold coffee, Orlena suggests a
chocolate brownie.
Customer orders pasta, Orlena recommends garlic
bread and a mocktail.
No staff involvement. No training required. Works
on any existing QR code setup.
THE RESULTS
Early results show a 37-40% increase in average
order value from week one.
For a cafe serving 80 tables a day at Rs 400 average
bill — that is over Rs 3.6 lakh in additional monthly
revenue from the exact same customers.
THE TECH STACK
Built on React + Vite frontend, Node.js backend,
PostgreSQL database, OpenAI GPT for the recommendation
engine, and Twilio for WhatsApp notifications. Hosted
on Vercel.
THE HONEST PART
We are early. We are actively onboarding cafes in
Bangalore right now. We do not have hundreds of
customers yet. But the problem is real, the solution
works, and the market in India is massive.
India's cafe and restaurant segment is estimated at
$18.83 billion in 2025 and projected to reach $30
billion by 2030. Urban Indians dine out 5 times a
month on average growing to 7-8 times. Every single
one of those visits is a missed upsell opportunity
without a system like this.
WHAT I AM LOOKING FOR
We just launched on Product Hunt today. I would love:
1. Honest feedback on the product and positioning
2. Advice from anyone who has sold to restaurants
or cafes before it is a notoriously hard market
3. Connections to restaurant owners or food tech
investors who might find this interesting
4. An upvote on Product Hunt if you think this
is worth building
Product Hunt: producthunt.com/posts/orlena
Website: hello.orlena.talk
Happy to answer any questions about the product,
the tech, the market, or the founder journey.
Thank you for reading.
Vishnu
Founder, Orlena
I like that you called out the QR menu being just a digital PDF, thats exactly where a lot of restaurant tools stall. Some teams stay on static menu pages or Uniqode, I'm building ScanLoop more around placement-level scan history so a cafe can tell which table card actually pulled the scan before they layer on upsells. If you can show both AOV lift and which QR spot created the order, the pitch gets way easier imo.
The insight isn't the AI it's that you watched 47 cafes and noticed the same pattern: servers aren't failing to upsell because they're bad, they're failing because they're busy. That's not a training problem. That's a capacity problem. The system doesn't replace the server. It gives them back time they didn't have.
I acknowledge and appreciate your perspective that it will give servers more time to do things that matter to increasing revenue per day.
Visiting 47 cafes before building is the ultimate masterclass in validation. Most founders (myself included sometimes!) tend to hide behind the code, but you actually went out and found a specific, measurable leak in their revenue.
That 37-40% increase in AOV is a massive hook. Selling to restaurants is notoriously hard because of the churn and thin margins, but showing them 'found money' from customers they already have is the best way to open that door. Great job on the PH launch!
Thank you , and yes you are right the 37 to 40 percentage is something that will happen after 3 or 4 months of orlena selling your cafe or restaurant items