In February 2025, I sat down with a blank page and started drawing services. Not writing code. Drawing boxes and lines. How an order would travel from a customer's phone to a kitchen screen. How a loyalty point would be earned, stored, and redeemed. How inventory would decrease every time a dish was served. This took months. The first commit came much later.

The first working version ran on Open Weaver Studio, a developer platform I helped build as one of the core team members. It was a simple QR ordering system. Not linked to anything. No kitchen display. No payments. No inventory. Just: customer scans, order goes somewhere, someone reads it.

Then I stopped. Other work, other commitments. For a few months Coda was just a folder on my laptop. I came back to it in pieces: a feature here, a week of nothing, three features in a weekend. That is what building a real system looks like from the inside. It is not a sprint. It accumulates.

I printed out the latest quarterly menu for Bridge. ₹8,000 to design. ₹12,000 to print and laminate. Three weeks later, tape covered three items and eight pieces of paper were stuck around the dining area.

A QR code seemed obvious. I built it that Saturday. Customers scanned it. The tape came off. The first-week complaints dropped to zero.

I should have stopped there. I did not.

The QR menu was a digital menu. Customers scanned, saw photos, read prices. But they still waved down a waiter to actually place the order.

That felt wrong. If you are already on your phone, you should be able to order from your phone. One tap, done.

Except now the kitchen needs to receive that order. The waiter app needs to see it. The system needs to log it, track it, and close it when the food arrives. A QR menu that can order is not just a menu anymore. It is a system. That is when the scope of what I was building changed entirely.

I have worked in Bridge's kitchen. I know what happens when a KOT slips under a prep counter during Friday lunch. You find out twenty minutes later when the table flags down a waiter and their food still has not arrived.

I was watching The Bear one evening: the expo board, the colour-coded tickets, Carmy calling out each order with the ticket in his hand. I kept thinking: why doesn't Bridge have that?

I built a kitchen display in two weeks. We stopped losing tickets.

People assumed we were building something to replace the waiter. That is not what this is.

Walk into any good restaurant and watch what a great waiter actually does. They remember you take no sugar. They notice the table that has been waiting before anyone waves. They defuse a difficult guest without anyone nearby noticing. No app does any of that.

What we built handles the parts that do not need that skill. The order goes from the customer's phone straight to the kitchen. The alert fires when the food is ready. The waiter is not spending twenty minutes chasing paper slips down a hallway.

We also built a full waiter app for guests who do not want to scan anything. Walk in, sit down, a waiter takes the order the old way. It all goes through the same system. The power stays with the person on the floor. We just gave them better tools.

PetPooja. ₹40,000 per year, base plan. Add-ons: ₹16,000–20,000 each. Our old system cost ₹10,000 a year, but it ran on local WiFi only. Router goes down, no orders.

I spoke to seven other restaurant owners across Palakkad and Thrissur. Same tools. Same complaints. Same workarounds. This was not a Bridge problem. It was every restaurant's problem.

Every morning at Bridge, someone sits with a stack of invoices. Twenty, sometimes thirty vendors. Ice cream from one supplier. Bread from another. Produce from the market. Meat from the cold store. Each invoice different. Each amount to be written somewhere.

The purchase manager tracks all of it: what arrived, what was paid, what is pending, what credit is owed. At the end of the month, reconciling that notebook against bank records is a day's work by itself.

I watched this for months. I did not add a feature for it. I built an entire module. Daily purchase entries, vendor ledger, credit and debit columns, receipt photos attached to each line. Not because it was on a product roadmap. Because I kept watching the people I know spend hours every week doing something that should take twenty minutes.

Oneiros Coda: brand strategy, campaigns, digital marketing. And I kept watching customer data accumulate inside our own system and do nothing.

Every regular who comes on weekday mornings. Every family that comes on birthdays. Every teenager who orders the same iced coffee every Saturday. Their data was there. Nobody was using it. A birthday message with 10% off costs almost nothing to send. It makes someone feel like they belong here.

Not daily active users. Not GMV. Whether someone walks into Bridge and feels like a regular. Whether the waiter knows they take oat milk. Whether their birthday gets acknowledged. Whether the chef can get to sleep before midnight.

That is what Coda is actually for.