Conviction before velocity
Speed without direction is expensive noise. We move when the thesis is sharp — not when the calendar demands theatre.
Ask why before how fast.
Open to a small number of conversations
Renato Xhaferaj
Build what
endures
A manifesto for engineers who care about craft — and investors who care about conviction. Written for the people who can tell the difference between a system built to last and a system built to be sold.
- 01
Judgment over volume
Fewer bets, argued properly, carried to the end.
- 02
Legible systems
Architecture a good team can still explain at midnight.
- 03
Reality, early
Bad news travels first here. That is the whole trust model.
[ 01 — Manifesto ]
Build what endures
A point of view on craft, capital, and the kind of work worth doing together.
I am not here to impress you with jargon. I am here because the best engineers and the best investors share one instinct: they can tell when something is built to last — and when it is built to be sold.
For several years, that instinct has guided how I work. Not as a performer on a stage, but as someone in the room when the hard calls get made: what to build first, what to refuse, when to move fast, and when to hold the line.
Most of what fails does not fail loudly. It fails quietly, months earlier, in a decision nobody wrote down — a shortcut taken to hit a date, a disagreement smoothed over instead of settled, a number reported the way it should have been rather than the way it was. By the time the failure is visible, it is only arithmetic.
The ventures I touch are meant to compound. That means respect for the people who write the code, respect for the people who allocate the capital, and zero tolerance for complexity that exists only to sound smart.
Great outcomes are rarely loud at the beginning. They are clear. Clarity is the whole strategy.
Five principles
The rules I do not negotiate — the ones that decide everything downstream.
Craft is leverage
Elegance in execution is not vanity. It reduces risk, attracts exceptional people, and tells investors you mean what you say.
Quality is a cost structure.
Complexity must be earned
Hard problems deserve hard thinking. Performative difficulty does not. The goal is a system a smart team can still explain at midnight.
Simple is the hard version.
Trust is built in quiet years
Reputation is not a landing page. It is what you do when nobody is applauding: how you treat partners, how you handle setbacks, how you report reality.
Character compounds too.
Capital follows clarity
Investors do not fund confusion — they fund narratives they can underwrite. Engineers do not stay for chaos — they stay for missions they can believe in.
One honest paragraph beats ten slides.
If this resonates, we should talk. If it does not, you will know quickly — and that is a kindness too.
[ 02 — The practice ]
Where I am useful
Three places where a second set of hands and an honest second opinion tend to change the outcome.
Product and architecture
Turning a vague ambition into a system with a spine: what the first version must prove, what it is allowed to ignore, and where the seams go so the second version does not require a rewrite.
- Scoping the first honest version
- Architecture that survives the next hire
- Killing the features that quietly cost the most
Execution and delivery
Shipping is a discipline, not a mood. Sequencing the work so progress is visible weekly, risk is retired early, and the team spends its energy on the parts that actually decide the outcome.
- Sequencing work against real risk
- Standards a team can hold without a policeman
- Cutting scope without cutting the point
Technical judgment for capital
Reading a technical story the way an operator reads it. What is genuinely defensible, what is a rebuildable weekend, and which claims in the deck the team could not survive being asked about twice.
- Separating moat from momentum
- Reading a team's real velocity
- Naming the risk nobody has written down
[ 03 — How I work ]
Plainly, and in that order
No discovery theatre. Four steps, each one designed to make the next one cheaper.
- 01
Listen properly
The first conversation is not a pitch in either direction. I want the constraint you have stopped mentioning because you assume it is permanent.
- 02
Frame the real problem
Most briefs describe a symptom. Written back in one paragraph, the actual decision usually turns out to be different — and smaller — than it looked.
- 03
Build the smallest honest thing
Something real, in front of real users or real numbers, early enough that being wrong is still cheap. Opinions are expensive; evidence is not.
- 04
Hand it over clean
Documented, legible, owned by your team — not by me. The measure of the work is how well it runs after I stop touching it.
[ 04 — Who this is for ]
Two rooms, one standard
Different seats at the table. The same bar for honesty, depth, and follow-through.
For engineers
- You want a builder who has been in production — not only in pitch decks.
- You care about judgment: what to automate, what to simplify, what to delete.
- You would rather ship something legible than something clever that owns you for years.
- You want disagreement handled in the open, on the merits, and then settled.
If you are tired of rewriting the same mistakes, we speak the same language.
For investors
- You want an operator who respects dilution, runway, and the cost of distraction.
- You look for teams that report reality early — not after the quarter is lost.
- You back people who can articulate the bet in one honest paragraph.
- You would rather hear the risk from me than discover it at the next board meeting.
If you allocate to substance over story, we should compare notes.
[ 05 — Contact ]
If the work aligns, reach out
I take a small number of conversations seriously: founders with a real wedge, engineers with a real problem, investors with a real thesis.
- I read everything. I answer what I can answer well.
- No form, no funnel, no mailing list — just an address.
- If it is not a fit, you will hear that quickly and plainly.
A few sentences on who you are and what you are building is enough. Noise is optional.
Usually a reply within a few days