The client is one of the best-known media companies in its market, part of a major international media group, and the operator of a leading OTT/VOD platform. The platform streams a broad catalog of series, films, and live TV to a mass audience, monetized through subscriptions and VOD packages.

In this business, revenue is inseparable from programming moments. When a flagship show premieres, interest doesn't rise gradually. It spikes, often within minutes, as a huge wave of viewers tries to subscribe or buy access at once. In the media industry this pattern has a name: the flash crowd, a near-instantaneous convergence of demand triggered by a single event. For a linear broadcaster, a flash crowd is simply high ratings. For a transactional streaming platform, it is something far more demanding, because every one of those viewers is trying to complete a purchase, receive an entitlement, and start watching in the same narrow window, and each one represents revenue that either lands or is lost in real time.

That dynamic makes the commerce layer mission-critical. It has to convert an enormous, momentary surge of demand into completed, correctly entitled purchases, in real time, on every device from a phone to a Smart TV. The platform's ability to capture revenue is only as good as its ability to stay up and transact when everyone arrives at the same second, which is precisely the moment when infrastructure is most likely to fail and when the business can least afford it.

Pretius designed and implemented a complete ecosystem composed of three integrated systems, an eCommerce platform, a central Product Catalog, and a dedicated Billing system, built on a microservices model and optimized for high horizontal scalability on AWS. Decomposing the platform into independent services was itself a deliberate answer to the flash-crowd problem: it let the components most exposed to traffic spikes scale horizontally and independently, rather than forcing the whole system to scale (and fail) as a single block. The central deliverable was a high-throughput sales engine integrated with multiple payment gateways, paired with a checkout path optimized for mobile devices and Smart TVs.

The engagement ran through a disciplined sequence of phases, from analysis and architecture design through development, performance testing, go-live, and ongoing maintenance, following a Waterfall approach in the initial build. That structure suited a project where the architecture had to be proven against load before launch, not iterated into shape afterwards: performance testing was a named milestone in its own right, not an afterthought. A stable seven-developer Pretius team delivered the core implementation, an effort estimated at roughly 700 person-days in the initial phase (about five months of work for the seven-person team). Since go-live, the collaboration has shifted to an Agile model of continuous development and maintenance, a mode it has stayed in ever since.

It's worth being explicit about scope, because the ecosystem's job was tightly defined. Video production and the transmission infrastructure (the CDN) were deliberately left out of scope. Pretius built the commerce and entitlement brain of the platform, the layer that turns demand into paid, authorized access, and integrated it with the streaming infrastructure that delivers the content itself.

A public, high-traffic sales platform is a standing target, and the project treated resilience as a first-class requirement rather than a hardening pass at the end. The team went in with an explicit risk assessment, and, tellingly, most of what could go wrong was anticipated and designed for in advance, with only the payment-gateway failures arriving as a genuine surprise.

The operator didn't need a generic online store. It needed a commerce ecosystem that could survive flash crowds, unlock purchased content in real time, and run smoothly on everything from phones to older Smart TVs. The "obvious" answer, a boxed platform, would have failed at exactly the requirements that mattered most: throughput at peak, deep integration with VOD entitlements, and device-level optimization for Smart TV. By choosing the harder path, three integrated systems on a scalable AWS microservices architecture, engineered for transactional integrity, entitlement-aware caching, and constrained devices, then hardened against DDoS attacks, fraud, and even payment-gateway outages, Pretius built a platform prepared for over 60 million page views a month and stable at the scale of the market's biggest premieres.

The result has been a key revenue channel for the operator, and a partnership Pretius has continued to grow over many years. For any transactional platform whose busiest moment and its most fragile moment are the same moment, that's the whole game: infrastructure that is most dependable precisely when everything depends on it.