Hotel Patliputra Continental · Hospitality
Hotel Reservation System
Direct bookings that clear, instead of failing at the payment step
- Role
- Full Stack Developer & IT Support
- Year
- 2021
- Stack
- PHP · MySQL · HTML5 · CSS3 · JavaScript · jQuery
The problem
An independent hotel taking bookings through a system it does not control pays for every reservation twice — once in commission, once in the guest relationship it never owns. The obstacle to running direct booking is trust: if the payment step fails, the guest does not retry, they go back to the aggregator. Inventory had to be accurate in real time, because a double-booking costs more than the reservation does.
The approach
Payment reliability was the whole engagement. I integrated CCAvenue, which Indian guests recognise and already trust, and built the booking flow so that inventory was reserved and released deterministically around the payment rather than optimistically before it. Database performance mattered for the same reason — a booking screen that stalls under load produces abandoned transactions, so the two problems are the same problem.
What I built
Real-time room booking and inventory
Availability held accurate across concurrent bookings, so the hotel could sell directly without risking double-allocation.
CCAvenue payment integration
A payment path guests recognise, wired so that a failure at the gateway releases the held inventory instead of stranding it. Successful transactions rose 38%.
Admin dashboard with booking analytics
Gave the front desk its own view of bookings and revenue rather than a spreadsheet reconstructed after the fact.
Database performance work
Improved database performance by 35%, letting the same infrastructure absorb three times the traffic.
Results
- Successful transactions
- +38%
- Booking efficiency
- +70%
- Traffic handled on the same infrastructure
- 3×
- Database performance
- +35%
How to read these
- This is a PHP and MySQL build from an earlier engagement. It is included because the outcome is strong and the problem is one hospitality clients still have — not as a sample of current stack choices.