What it looks like when research lands

Designing public policy from the territory, not the desk.

The state designs policy from behind a desk. It decides what people need without asking them, implements, and only then discovers it didn't work. The problem wasn't a lack of will — it was that nobody inside the bureaucracy knew how to do it any other way. Schools of government taught management, not design.

I created and led the Public Policy Design Academy of Argentina's national government. Teams came in with a real policy and left with that policy redesigned: ethnography and service design applied to their own work while they studied.

Apolitical ranked it the second-best government academy in the world, alongside the UK's GDS.

Redesign the relationship, not the paperwork.

In 2016, 12% of residents in Palpalá paid their taxes. Without revenue there is no public works, and without public works nobody sees a reason to pay. The municipality was inside that loop and read it as a collections problem. It wasn't.

I created and led the Innovative Cities Ecosystem, which brought public innovation methods to 60 Argentine cities. In Palpalá, the municipal team went out to understand why people weren't paying: what stood in their way, what motivated them, what they expected in return. The answer wasn't in the tax system — it was in distrust. Residents didn't know where their money went. So the team took payment into the neighborhoods, opened the books, and let residents decide which public works the money would fund.

Collection went from 12% to 60%. With that, the municipality could fund its own construction.

Accelerating reskilling across 200+ manufacturing plants

When a company announces it will automate its plants, the problem gets framed as technical: which tasks get replaced, which courses to run. But an automation process isn't solved by teaching. It's solved by understanding the behaviour of every actor who has to move for the transformation to happen.

I was Design Lead and author of the method. Working with the global manufacturing HR team of one of the world's largest consumer goods companies, we researched how the four actors in the system experienced the change: the factory workers, who have to open up to a process of transformation and learning, understanding that their tasks are shifting into the future; the union, which protects what exists; the line managers, who translate between leadership and the shop floor; and the HR leaders, who hold the process from above.

Each resists differently, and each moves differently. The method designs a specific intervention for each actor, at different levels of maturity, with frameworks and tools so that every factory can run its own reconversion process.

The point was for every worker to be the protagonist of their own future.

One hundred hours inside the camp.

Risk management on an oil field belongs to engineering: protocols, indicators, audits. But there's a layer the protocol can't reach — why people do what they do when no one is watching. That layer is where risk is actually decided, and it appears on no dashboard.

I led the first ethnographic immersion in the oil industry on risk perception. One hundred hours of participant observation inside the camp, in-depth interviews, and surveys. The result was a body of findings about operational culture that the company had no instrument to see: the reasons behind the numbers it already had.

With those findings, the company built its action plan for the operational culture of the field.

Fourteen thousand skills, one possible path.

Public training was designed in a straight line: defined at the top, delivered at the bottom, with no mechanism to feed back what the market was actually asking for. The result was free training that didn't help anyone move jobs.

I was Product Lead of Relink and led the full technology team, from concept to production. The work began off-screen: interviews, workshops, and validation sessions with companies and subject-matter experts to name the skills industry was really looking for. My team built the job profiles and, on top of them, a taxonomy — an adaptation of the European Commission's ESCO model to the Latin American context — and an algorithm that calculates the gap between the role a person holds and the one they could hold, returning a viable learning path.

Funded by the IDB and Sofofa.

Live and in production at relink.cl · The taxonomy is open at MAPHA

The map is drawn by whoever walks it.

Learning pathways in the public sector are almost always designed in a straight line: a catalogue of courses, a suggested order, an expected rate of progress. The assumption underneath is that the problem is one of supply.

It was not a supply problem. Nobody knew what skills were already inside. Without that map, any capability plan is a bet — and in a sector where roles are being reconverted, betting is expensive.

I worked as implementation advisor alongside a global research team, bringing in what I had learned from earlier labour reconversion work. The principle we held to: the diagnosis cannot be an external assessment. A civil servant will not complete a tool that audits them. They will complete one that serves them.

The Tracker was built around two tracks — data and gender — chosen not as themes, but as the two capabilities where the gap between what the state needs and what its people hold was widest.

Research-backed, designed with civil servants, and funded by the Gates Foundation.

Live and open at Apolitical