How SRPRS.me automated 80% of its bookings and removed the operations bottleneck
- 80%
- Bookings completed without manual intervention
- 3
- Booking categories replacing one manual process
- GREEN / ORANGE / RED
- Rules-based booking logic
The Client — SRPRS.me
The clientSRPRS.me is a surprise travel platform. Customers pick their travel dates and preferences — flights, hotel, and destination are revealed only shortly before departure. The booking process is not a back-office task. It is the product.
If the booking process breaks or slows, revenue slows with it. There is no margin for a system that cannot keep up.
DWS has built and maintained the Django platform behind SRPRS.me for several years. This is not a project relationship. DWS co-invests in the platform alongside the SRPRS.me team.
The Problem
The problemEvery booking used to require a person.
The operations team checked availability, compared prices, applied business rules, and confirmed bookings manually. For every single trip.
It worked when volume was low. As the platform grew, the team’s capacity became the ceiling on how many bookings could complete in a day.
The process was slow. It was error-prone. And it could not scale. Hiring more operations staff was not a product strategy.
A booking system that handles the routine, so the team can handle the exceptions.
DWS built an automated booking system on top of the SRPRS.me Django platform. It processes 80% of all trip bookings from submission to confirmation without any human step in between.
The system uses rules-based logic to categorise every incoming booking:
- Green:the booking meets all criteria — auto-book immediately
- Orange:one or more conditions need review — flag for the operations team
- Red:outside the rules — manual handling required
The logic accounts for destination, margin, travel dates, and live availability. Getting availability right in real time required integrating with the Travellanda API — and building a custom caching layer to handle the volume without hammering the external API on every request.
Django background tasks handle eligibility checks and execute bookings automatically. The operations team manages exceptions. Not the routine.
The platform scales without the team scaling with it.
80% of bookings now complete without manual intervention. The system runs on the same Django platform — no new infrastructure, no separate tool, no third-party automation layer.
The operations team went from managing individual bookings to handling only the edge cases that genuinely need a person.
The capacity freed up went into improving the platform, not managing the backlog.
When booking volume grows, the automation handles it. The team does not.
“Integrating with the Travellanda API and building our own caching system for real-time price availability were the hardest problems to solve. We pushed through them. The result cleaned up operations and opened the door for the platform to grow.”
A challenge like this in your platform?
Manual processes that cannot keep up with volume are an engineering problem, not a hiring one. The fastest way to find out if we can solve yours is a 30-minute call.