Antonio Corzo Rico

An area organized in silos, working as a chain and managed through spreadsheets, where everyone saw their own part but nobody owned the outcome.

Starting point

Work fragmented and constantly handed off between functions, with no product view and no operational metrics. Invisible bottlenecks and frequent rework.

Intervention

Redefining the unit of value: from scattered tasks to the course as a complete product. Validated with small cross-functional pilot teams and then scaled to six teams with end-to-end capability, their own metrics and a defined working cadence.

Impact

~50% less production time and ~80% fewer structural dependencies in the pilot, with a marked drop in rework and better quality. The area began managing itself without constant supervision, and growth stopped meaning added complexity: it now means replicating teams that already work.

An area absorbing every incoming request without filter or prioritization, where decisions were driven by pressure rather than data.

Starting point

Chronic overload, excessive multitasking and requests left forgotten indefinitely. Everything was urgent and there was no visibility of the real state of anything.

Intervention

Redesign of the request value chain, with a validation step that improves data quality before work starts, and an explicit limit on work in progress. Reorganization into four specialist teams plus a data administration team, with dashboards and expiry criteria for stale work.

Impact

Shorter delivery times and less multitasking, better workload distribution and no more forgotten requests. For the first time the area can measure its real capacity, size its resources and decide with data instead of perceived urgency.

A large area with several intake channels and much of the day-to-day work split between the internal team and an external supplier.

Starting point

Requests arriving by email, chat and meetings with no common filter, and work bouncing back and forth between teams. With no shared view, it was hard to tell what each side was working on at any given moment.

Intervention

Progressive centralization of intake, a prioritized backlog, a work-in-progress limit and a clear owner for prioritization. A shared workspace with the supplier, a formal validation state for deliverables and joint weekly planning.

Impact

The area took over deciding what comes in and in what order. The shared workspace gives both sides the same view of work in progress and of what has been delivered, which simplifies joint planning and puts the supplier relationship on a clearer footing.

Two areas that depend on each other but planned, prioritized and measured differently, with a framework that looked sophisticated on paper and was hard to govern in practice.

Starting point

Work was fragmented by team, which meant constant coordination just to understand what value was being delivered. Exploration and execution lived in the same place, more initiatives were opened than closed, and there was no established link between annual objectives and day-to-day work.

Intervention

One synchronized process for both areas. The unit of value became self-contained, even when several teams are involved, so the value explains itself. Exploration was formally separated from delivery to stop the two competing for the same capacity. Simultaneous work was capped at three active fronts per team with weekly prioritization, and every piece of work was tied to a key result in the strategic plan. A regular review with the leadership committee was added.

Impact

Product and Technology now share one mission, one view and one cadence. Strategy is traceable through to execution: leadership can see what is being worked on and which objective it serves. Less fragmentation and fewer cross-dependencies mean better estimates and more reliable dates, and teams can defend their priorities with data instead of absorbing everything that comes in.

The connection between business and technology — requirements, stories, roadmaps, reviews — was handcrafted: it consumed senior capacity, produced errors and was a permanent bottleneck.

Starting point

The work of translating what the business asks for into what teams build depended on specific people, with waiting and context loss at every step. Team performance was evaluated on perception.

Intervention

Generative AI embedded at specific points of the flow, always on top of processes that had already been redesigned. In discovery: co-creation of personas, user story mapping and adaptive roadmaps. In the link between business and technology: AI-assisted stories and initiatives, prototypes and landing pages. In management: team performance reviews and impact reporting automated with parameterizable dashboards.

Impact

From days to hours for product definition and planning, and refinement stopped eating whole team sessions. Landing pages, prototypes and proofs of concept in hours, with no dependency on other areas, where there used to be days or weeks of waiting: a hypothesis is now validated in the time it used to take to request it. Quarterly and yearly impact reporting went from handcrafted to instantly available, and performance evaluation runs on continuous data rather than perception.

Two areas that depended on each other but worked to a traditional model with no connection between them. The oldest case on this list, and the one that explains why the rest work.

Starting point

Technology and product each planned, prioritized and measured on their own, on an inherited way of working. The dependency between them was real, but there was no shared framework to manage it.

Intervention

Two years of work across both areas: a shared framework for planning, prioritization and delivery, and continuous delivery in short cycles instead of large batches. The piece that changed things most was defining the product jointly across both areas, with Agile Inception and, above all, the MVP: no longer negotiating the full scope up front, but agreeing on the smallest version that solves the problem. And, as part of the same work rather than as a consequence of it, the intervention on culture: ownership of the outcome, transparency and continuous improvement.

Impact

Dates became reliable and delays dropped, and with that stakeholder expectations became manageable: instead of one distant date to defend, there was progress to show every few weeks. The change held because the culture changed with it, and that is the lesson that orders everything else: a framework rolled out without culture change is dropped the moment urgency bites. Years later I met the same problem at Northius, at greater scale, and solved it in the same order.