How I Use AI to Run 4 Restaurant Locations

The Reality of Multi-Location F&B

Running one restaurant is hard. Running four locations across two brands — Byond Walls Pizza Bar (2 outlets) and Songhwa Korean Cuisine — is a logistical puzzle that never stops.

Stock moves between shops and warehouses. Kitchen and bar track ingredients differently. Staff rotate between locations. Purchase orders need approval chains. And every morning, I need to know the state of everything before I can make decisions.

Most restaurant owners solve this with WhatsApp groups, spreadsheets, and gut feeling. I solved it with AI.

Three Bots, Three Roles

I built three AI assistants on Telegram — each with a specific job:

Flash is the quick responder. Any staff member can message Flash and ask: "How much mozzarella at SS2?" or "What orders are pending?" Flash pulls the answer from our live inventory system and replies in seconds. No need to call a manager, no need to open any app.

Kimi handles operational questions — shift details, kitchen workflows, standard procedures. Think of Kimi as the team's always-available operations manual.

CEO is my personal briefer. Every morning at 8am, CEO sends me a complete business snapshot: stock levels across all locations, pending supply orders, petty cash spend, items approaching expiry, and a priority action list. Two minutes of reading replaces 45 minutes of spreadsheet-checking.

What the Bots Actually See

The bots connect directly to our custom inventory system via API. They have real-time access to:

• Stock levels across all 4 locations (2 shops + 2 warehouses)
• Supply orders — requested, approved, received, with full audit trail
• Petty cash — daily expenses across locations
• Product freshness — how long items have been sitting at each location
• Staff schedules — who's working where and when
• Website traffic — how byondwalls.com is performing

The Impact on Daily Operations

Before AI: I'd start every morning opening 3-4 different apps, scrolling through WhatsApp messages from managers, and trying to piece together what happened overnight.

After AI: I open Telegram, read the CEO brief in 2 minutes, and know exactly where every business stands. If something needs my attention, it's highlighted. If everything is fine, I can focus on strategy instead of firefighting.

My staff stopped depending on managers for basic information. They message the bot and get answers instantly — stock counts, order statuses, anything they need. This removes bottlenecks and speeds up decision-making at every level.

Why Telegram?

Every restaurant worker in Malaysia already has Telegram on their phone. There's no app to download, no training needed, no login to remember. They just message the bot like they'd message a colleague.

That adoption advantage is massive. The best system in the world is useless if people don't use it.

Building It Yourself vs. Buying

Off-the-shelf restaurant AI tools exist, but they're generic. They don't understand that my kitchen tracks stock differently from my bar. They don't know that I need alerts grouped by outlet. They can't pull from a custom inventory system built specifically for how my restaurants operate.

Building my own AI layer took time, but the result is a system that fits like a glove. Every bot response, every alert threshold, every data point is tailored to how I actually run my business.

What's Next

The bots are getting smarter. I'm working on predictive stock alerts ("You'll run out of pizza dough at SS2 by Thursday based on current usage") and automated reorder suggestions. The goal is a system that doesn't just report — it recommends.

If you're running multiple F&B locations and drowning in operational noise, AI isn't a luxury. It's the difference between managing reactively and managing proactively.

Want to see how this works in practice? Check out the [full case study](/case-studies) or [get in touch](/connect).