Travsta started with a problem we kept seeing inside the travel industry.
Many travel companies were still running a large part of their operations through Word documents and Excel sheets. Itineraries were manually assembled, hotel rates lived in spreadsheets, costing happened separately, content had to be copied repeatedly, and every customer change created another round of work. In some companies, producing a single itinerary could take around three hours.
In 2022, Shazeen and I started Travsta inside Purple Software because we thought there had to be a better way.
The original idea was simple: take everything required to create and sell a tour and put it into one system.
The first version didn’t work
We built the first version of Travsta believing we understood the problem.
We didn’t.
Technically, the software worked. But once we put it in front of actual travel companies and watched them try to use it in their day-to-day operations, the gaps became obvious.
The travel industry has an enormous number of small operational details that are difficult to understand from the outside. Hotel contracts have seasons, room categories, meal plans and different pricing structures. Tours can be B2B or B2C. Different agents need different information. Costs can involve multiple currencies. Customers constantly request modifications. Each company also has its own way of presenting and selling a trip.
Our first version had not accounted for enough of this complexity.
So we kept changing it.
Travsta went through thousands of iterations across workflows, interfaces, costing logic, content management, itinerary presentation and rate management. It took roughly two years of continuous development before I felt we had something that travel companies could genuinely depend on in real-world operations.
That process taught me one of the biggest lessons I’ve learned about building products.
Don’t build the SaaS first
There is a temptation when building a software product to create the platform, put a signup button on a website and immediately try to sell subscriptions.
I think that’s usually backwards.
Before turning something into SaaS, the product should survive inside a real company.
A real customer will expose problems that a product team will never discover sitting in a meeting room. They will use the system differently than expected, encounter edge cases you didn’t anticipate and complain about seemingly tiny things that turn out to be critical to their workflow.
We learned far more by working closely with travel companies than we ever could have learned from simply designing what we thought the industry needed.
Only after repeatedly testing Travsta with real users and improving it around their operations did it begin to make sense as a product that any travel company could sign up for.
That lesson now influences how I think about almost every SaaS product: prove the workflow with real customers first, productize it second.
Why are we still sending PDFs?
Another thing Travsta made us question was how travel companies present their products to customers.
PDF itineraries made sense years ago. In 2026, sending someone a large PDF and expecting them to zoom around it on a phone feels outdated.
Most customers are opening links from WhatsApp, email or another mobile application. The itinerary should behave like the rest of the internet.
So Travsta allows companies to send customers a web link instead.
The itinerary automatically adapts to the customer’s device, making it much easier to browse the trip, see destinations, accommodation, experiences, photographs, maps and daily plans. Companies can use their own custom domain, while Travsta can also notify the travel company when a customer views the itinerary so the sales team knows when to follow up.
PDF and Word exports still exist because businesses need them, but they are no longer the only way to present a trip.
What Travsta has become
What started as an itinerary builder has gradually become a much broader operating system for creating and selling tours.
Travel companies can maintain reusable libraries of destinations, experiences, inclusions, exclusions and other content instead of recreating them for every inquiry. AI through TravGenAI can help generate destination content, itinerary descriptions and day highlights.
Tour costing is handled directly inside the platform, including support for multiple currencies and different costing structures. Companies can also manage hotel contract rates, seasonal pricing, room categories, room types and meal plans. Travsta can even extract hotel rates from uploaded contract documents using AI, after which the user reviews and approves the extracted information.
The platform now includes itinerary templates, multilingual itineraries, maps and travel routes, B2B agent collaboration, customer and agent management, itinerary status tracking, client-view notifications and branded sharing through custom domains.
The goal behind all of those features is still the same as it was in 2022: remove as much repetitive work as possible between receiving an inquiry and sending the customer something they can buy.
From three hours to minutes
The original benchmark that bothered us was the roughly three hours some travel companies were spending to prepare an itinerary.
Today, Travsta is designed around creating polished, client-ready itineraries in minutes rather than hours. That is still how we measure the usefulness of the product.
Travsta is now being used by travel professionals around the world, including travel agencies, tour operators, DMCs and travel designers. The product has moved well beyond the first version we built inside Purple and continues to evolve through feedback from the people actually using it.
Looking at the changelog today is almost a record of that process. There have been continuous additions and refinements covering AI generation, costing, hotel contracts, agent collaboration, multilingual content, templates, previews, exports, maps and dozens of smaller workflow improvements.
There was no single breakthrough feature that made Travsta work.
It was thousands of small corrections.
What Travsta taught me
Travsta changed how I think about product development.
The difficult part isn’t building the first version. Modern engineering tools make that increasingly easy. The difficult part is building something that survives reality.
You have to watch people use it, discover where your assumptions were wrong, fix those problems and repeat that process until the software starts fitting naturally into how the business operates. Our first version of Travsta failed that test. I’m glad it did. Without those failures, we probably would have built a cleaner-looking product that nobody actually wanted to use. Instead, we spent years learning the problem properly.
And that eventually turned a three-hour manual process involving Word documents, Excel sheets and PDFs into something travel companies around the world can now do in minutes.