2026
Depth first, then simple
You cannot simplify what you do not deeply understand. Anything that looks effortless has somebody’s obsession underneath it.
renepupala®
LondonKošiceHong Kong
- AI engineering and agentic systems
- Product, from nothing to live
- Consulting and enablement
- Design discernment and taste
Where the work landed
(2019–26©)
- A global luxury brandAI lead, global transformation
- A global tech giantCustomer experience, two years
- A national construction companyConsulting
- EricFounder
- SalahFlowCo-founder
- FrameCoachCo-founder, paused
(4)
Ventures.
Four things I built. Two are live, one is paused, one was twenty-four hours. Each one is here for what it proved, not for how it looked.
- A property agent that lives in WhatsApp
- A prayer tracker for reverts
- An AI camera coach
- User interviews into shipped code
What I do
Practice.
(4)
Agents built on primitives and tools rather than a fixed decision flow, with a way to measure whether a change made it better before it ships.
Deciding what should exist at all, then building it. I have taken a product from a weekend to real users in production on my own.
Introducing AI to a programme from nothing, then running the enablement for senior people until the sceptical rooms are not sceptical any more.
If the code writes itself, the scarce thing is deciding what should exist and telling whether what came out is any good. That is the bottleneck now.
Positions I hold
(5)
My five takes.
The ones people push back on, written plainly rather than hedged. Held, not performed.
Software engineering is largely solved. If it is not completely solved it is largely solved, and the rest closes in months, not years.
You have to go deep before you can go simple. The depth is what makes the simplicity possible.
What is left for us is discernment, systems and taste. Those are not soft skills wearing a costume, they are the bottleneck.
People skills are worth more than ever, not less. The better the tools get, the more the difference comes down to the person holding them.
Good design with AI is not just possible. The constraint was never the tool, it was the assumptions the person brought to it.
- £30k
Partnership brokered across six companies
- 900+
Film school alumni in the FrameCoach beta
- 3
Dubai brokerages in commercial conversation
- 1st
Class, Computing Science, and an exchange year
rene pupala®
Go deep, then make it simple enough to hand over. I build the thing, I measure whether it actually got better, and I say plainly what happened. Including the parts that did not work.
2026
The corporate game is political
Nobody explains how decisions actually get made. It is a genuine skill, rarely taught, and constantly rewarded.
rene pupala®
What is left
for us?