Case study · Product development

VOD WFM – Video Infrastructure Built to Last

Millions of households worldwide watch video-on-demand content delivered through a platform we co-built with TiVO (now XPERI). It runs in 18 markets, processes 66 million tasks a year, and has been operating at production scale since 2016.

INDUSTRY Video / TV infrastructure · ENGAGEMENT 10 years · STACK Java · High-availability process engine · Multi-system integration

30+Users & households served
18Markets deployed
66MTasks processed annually
3M+VOD bundles processed

We built it with TiVO over four years (2016-2020). For the six years since, their own engineering team has owned and scaled it independently – and across that whole stretch, we’ve kept partnering with XPERI on other initiatives. Three things are worth keeping separate here: the build, the handoff, and the relationship that outlasted both.

Here’s what we built, what it changed for TiVO’s business, and how we structured the engagement.

Client

TiVO (a part of XPERI)

Project

VOD workflow manager – automated provisioning for operator infrastructure

Grew to

18 markets,
66M tasks a year

The problem

Engineering at the speed of TiVO’s market opportunity

TiVO had a clear go-to-market opportunity: provide cable operators with a video infrastructure package, including a Content Management System and automated VOD provisioning. The CMS opened doors. The VOD layer would scale the relationships.

The challenge was timing.

Building VOD provisioning meant:

  • A process engine with high availability and failure tolerance – operators can’t tolerate downtime that interrupts subscriber service
  • Complex multi-system integration – source film providers, metadata systems, encoding pipelines, and operator infrastructure all had to interoperate cleanly
  • A Java engineering team – strategically chosen by TiVO, but not yet built up internally

Standing up that team and shipping the platform in parallel – while the market opportunity was open – wasn’t feasible. TiVO needed a partner who could bring the engineering capability immediately, build the platform alongside their existing teams, and structure the work so TiVO’s organization would be positioned to operate it long-term.

We started that engagement in March 2016.

VOD WFM in numbers

Users and households served30M+
Localities deployed18
Individual tasks processed annually66M
VOD bundles successfully processed3M+
Years of collaborative development5
Years TiVO has operated it independently5 and going

What changed for TiVO’s business

The VOD workflow manager started as a supporting layer for TiVO’s mid-market operator package. As its capabilities matured, four things changed:

Enterprise operators became viable customers.

The modular architecture let TiVO adapt the same engine to each operator’s infrastructure without forking codebases. What started as a mid-market tool became infrastructure trusted by enterprise-scale operators worldwide.

It became a strategic product asset

As TiVO’s product strategy evolved, the CMS that was originally the lead product was sunset. The VOD workflow manager, by contrast, became a core component of TiVO’s offering for the world’s largest operators – a position it holds today.

New markets opened up

The platform’s design enabled deployment across 18 markets, each with its own operator requirements, regulatory environment, and content sources.

TiVO built an independent engineering capability

Through four years of co-development, their team built the institutional knowledge to operate and evolve the platform on their own. They took full ownership in 2020 and have scaled it under their own stewardship for the six years since.

The original brief was “build us this platform.” The outcome was a platform, a strategic product asset, and an engineering team confident enough to own it – and XPERI keeps coming back to us for other work, which is the strongest endorsement of how the first engagement was structured.

Process – from tight deadline to clean handoff

We didn’t build this in neat phases. We built it in stages, each one driven by what the operator needed next – and from the start, with an eye on the day they’d run it themselves.

March 2016 – November 2016

Kickoff under pressure

The first version had to ship for a Latin American operator on an aggressively tight deadline – tight given the scope on the table. We delivered the first working VOD workflow manager into production.

January 2017 – 2018

Hardening and reach

With the first release live, we cleaned up, broadened encoder support, and added multi-localization: the same content provisioned to several markets at once, since the operator ran the service across multiple countries.

2018 – 2019

The second-generation platform

 A rebuilt, properly modular architecture – per-client configuration without touching the codebase. New clients came on, along with integrations into encoding, content management, and CDN systems.

2019 – May 2020

Build it, then hand it over

More of the same evolution, running in parallel with something deliberate: building up the client’s own team to take the platform over. By the time we stepped back, they were ready to run it without us.

Four years. Built under pressure, evolved through new clients and markets, and handed off clean – and we’ve kept working with XPERI on other initiatives since.

Key features

The platform’s design reflected three commercial realities: operators couldn’t tolerate downtime, every operator needed something slightly different, and TiVO would eventually need to maintain it without us.

Multi-source content ingestion

Handles film files and metadata from multiple providers, normalizing them into operator-ready packages

High-throughput processing

Built for the 66M-task-per-year load profile, with headroom for growth

Java-native architecture

Chosen to align with TiVO’s long-term engineering hiring strategy

Documentation built alongside code

Architecture decisions, integration maps, and runbooks designed for engineers joining the project years after delivery

Have a similar problem? Let’s talk.

If your team has a complex technical challenge – scaling a methodology, modernizing a legacy system, or building a data product – we’d like to hear about it.

No deck, no sales pitch – just a conversation about what’s broken and what it would take to fix it.