Notes on building ventures that scale

Engineering Strategy

Choosing a tech stack for an early-stage startup: a framework, not a trend list

The stack decision founders agonise over rarely matters as much as the reversibility decision they skip. A practical framework for picking technology before product-market fit.

Read Founders

Choosing a business model before you build the product

The revenue model decides what the product has to be good at. Picking it late means reworking parts that are already load-bearing. A pre-build framework.

Read Venture Engineering

What is venture engineering, and how is it different from an agency?

Agencies deliver a scope. A venture engineering partner is accountable for whether the thing scales. Here is the practical difference.

Read Founders

In-house team vs. a venture engineering partner: how to decide

Hiring engineers is the obvious move. It is also slow, expensive to reverse, and risky before product-market fit. A framework for the trade-off.

Read Architecture

A pre-build architecture checklist for founders

Ten questions to answer before the first sprint. Getting these right is cheap now and expensive later.

Read AI Engineering

Shipping AI agents to production: what survives contact with real users

The demo is easy. The production system that stays accurate, cheap, and safe under real traffic is the actual project. What changes between the two.

Read Platform

Build vs. buy for internal platforms: a decision framework

Every internal tool starts as a spreadsheet and a good intention. A simple test for when to keep buying, when to build, and when to do neither.

Read Investors

Technical due diligence: what a serious review actually covers

A code walkthrough is not diligence. What an investor or acquirer should actually examine before wiring the money.

Read Infrastructure

Multi-cloud vs. single-cloud: a cost and risk framework

Multi-cloud sounds like insurance. It is also a tax on every engineer, every day. When the premium is worth paying.

Read Enterprise

Modernising a legacy system without a big-bang rewrite

The full rewrite is the most tempting and most dangerous option on the table. How to modernise incrementally while the business keeps running.

Read Founders

How to scope an MVP so it doesn't become a rewrite

A thrown-away prototype is fine. An MVP you have to throw away after it works is a failure of scoping. How to draw the line.

Read FinOps

Cloud cost optimisation: where the money actually goes

Most cloud bills are not big because of one wasteful thing. They are big because of ten small ones nobody owns. A practical order of attack.

Read Platform Engineering

What shared hosting can and can't do for a growing product

Shared hosting gets dismissed as a toy environment. It can run a real, production-grade platform — if the architecture is built around what it actually constrains.

Read