Sanará / Sp. “to heal, to be made whole.”

Huw Bevan.

I’m Huw. Fifteen years operating inside high-growth ventures, six of them at Uber, before becoming a founder in my own right.

Sanara is what I do now: operational transformation for PE-backed, scaling and later-stage businesses, working through their people, their teams, and AI. We assess how the business runs, then help you to improve it.

How I came to do this work.

After spells in investment banking and energy strategy & M&A, I joined Uber in 2015 as a city launcher: literal boots-on-the-ground work, opening new markets across the UK and Ireland. Over the next six years I moved through marketplace, central operations, and the UK leadership team.

In 2017 Transport for London refused to renew Uber’s licence. I built the UK Safety team from scratch in the months that followed, rebuilding relationships with TfL and the Metropolitan Police, engaging MPs, NGOs and public officials, and helping secure the licence renewal that kept 80,000 drivers in work. It is still the hardest, most clarifying work of my career.

I then ran the Rider business across UKI, launched new modalities, and led COVID response and recovery. I left in 2021 to become COO and later CEO of Gyre, a psychology-AI startup applying behavioural science to team performance, and to co-found Koyo Healthtech, a primary care platform live in Nigeria, where I am now an advisor.

What good operating actually means.

Strategy is won in the operating detail.

Plans are easy. The meeting cadence, who owns which metric, how decisions actually get made: that detail is where strategy holds or falls apart.

AI is an operating discipline, not a feature.

The companies that win with AI treat it as a way of working: chained, evaluated, supervised, not as a product spec. I build that discipline.

Change holds at the level of the team.

Transformation fails when it stops at the org chart. What decides whether change holds is whether a team actually works differently day to day: how it makes decisions, what it surfaces instead of avoids.

Distributed leadership is the only kind that scales.

If decisions only happen when the founder is in the room, the company has a ceiling. The work is to push real authority outward without losing coherence.

Less ceremony. More rhythm.

Most operating systems collapse under their own ceremony. The best ones are quiet, repeated, and easy for the team to keep running.

Who this is for.

I work best with PE-backed and later-stage businesses that know they need AI in the operating model but not where to start, and that are often preparing for a step-change, transition or exit. The common thread is a leadership team that wants to build the capability properly and make it stick, not just buy a tool and hope.

A leadership team’s job is to design how the company runs, not make every call inside it.

Most of my work comes back to one idea. The people closest to the work should be able to make the call, so decision rights belong as close to the information as possible. The leadership team builds the system that lets that happen: what matters most, who owns what, how decisions actually get made. Then it holds its nerve and doesn’t step in the first time things get uncomfortable.

The best teams I’ve worked with aren’t the ones that work hardest. They disagree well, change their minds when the facts change, and stay clear enough on where they’re going to say no to everything else. Most of what I do with founders and leadership teams is help them get there.

Start with a conversation.

30 minutes, free, no obligation. We talk about what’s actually in front of you, and whether I’m the right person to help. If I’m not, I’ll point you somewhere better.

[email protected] →