About
About Shelly Luciano
Brazilian-born, London-based. I build the commercial and operating machine that lets AI companies scale. Three promotions in five years at Leah, a SoftBank and Insight Partners-backed enterprise AI company. London Business School MBA. Strategy& before that.
My career looks like four different jobs and is really one. Engineering taught me that a plan and what happens are different things. Consulting taught me to find the real question. Product taught me what AI can actually do. AI at scale taught me that adoption is an operating problem.
I am most interested in the gap between what AI does in a demo and what it does inside a customer's operating model six months later. I call it the Execution Gap. Closing it is the work.
Four stages, one thesis
01
Rio de Janeiro, operating discipline
Graduate programme at Metrô Rio, then engineering and project management at Technip. Infrastructure at that scale teaches you quickly that the plan is not the thing. Stakeholders, constraints and consequences are the thing. I have never stopped thinking that way.
02
London Business School and Strategy&
MBA, then four years in strategy consulting leading teams across ten projects and six sectors. This is where problem framing became the core skill. Not answering the question in the brief. Finding the question that was actually being asked.
03
Monolith AI, the product side of AI
First product hire at an early-stage AI company, later acquired by CoreWeave. Product-market fit, roadmap, positioning, and the discovery work that comes before any of it. This is where I learned what AI capability actually is, and how far that is from a customer using it.
04
Leah, AI at enterprise scale
Joined as the AI wave broke and progressed to VP Strategy and Operations, three promotions in five years, leading a team across EMEA and LATAM. Built the operating cadence, forecasting and commercial infrastructure the company scaled on. Advised over 200 enterprise clients across the AI GTM lifecycle. Ran a 300-demo commercialisation programme. Created the Innovation Programme and ran more than 40 AI trials end to end with real client data in production-like environments. Currently redesigning the end-to-end post-signature customer journey around AI agents. Not tool implementation. Operating model redesign.
In one line
“A strategy that exists and a strategy that lives are different things.”
Most AI companies can prove capability. Very few can make it land inside a customer's operating model. Closing that distance is the work, and it is what I have spent the last five years doing at enterprise scale.
Selected timeline
- Metrô Rio
- Graduate programme, then senior project analyst, infrastructure portfolio
- Technip
- Project management and engineering, oil and gas
- London Business School
- MBA
- Strategy&
- Engagement Manager, ten projects across six sectors
- Monolith AI
- First product hire, acquired by CoreWeave
- Leah, formerly ContractPodAi
- VP Strategy and Operations, three promotions in five years
- Building
- Ship small products. Most recently a tool helping professionals manage their professional reputation using their LinkedIn data
On the record
Selected media
Transforming Strategy into Real-World Execution
Win Win Podcast, Episode 149, with Highspot
Closing the GTM Execution Gap
The GTM Gap Series, fireside with Lucas Welch, Highspot
Point of view
Three things I keep coming back to
i.
The Execution Gap is the real constraint, not capability.
Models are ahead of organisations and the distance is growing. Adoption needs trust, workflow fit, governance and a value event someone in finance will recognise. The companies that win this cycle will be the ones honest about that distinction rather than the ones with the best demo.
ii.
Agents change the operating model, not just the workflow.
Putting an agent into a process that was designed for people does not work. The handoffs, the escalation paths, the points where human judgment has to stay: all of it has to be rebuilt. Most organisations are automating steps. The work is redesigning the system.
iii.
Invisible cognitive load is a design problem AI should solve, not worsen.
The mental load carried by working families, disproportionately by women, is real, measurable and largely undesigned for. AI that adds to it is not neutral. It is a choice someone made.
I live in London with my husband Anthony, a primary school teacher, and our two children. Most of what I think about professionally is shaped by the household we are running and the world they will work in.