Build
How we build companies.
A repeatable path from problem to operating business — with explicit gates to continue, stop or scale.
- 01
Discover
Start from a real problem and a specific customer — never from a trend.
- 02
Validate
Prove demand and a revenue mechanism before we write serious code.
- 03
Build
Ship an MVP on shared, proven foundations with clear acceptance criteria.
- 04
Launch
Go live under its own brand, domain and customer relationship.
- 05
Operate
Run it with real metrics, its own P&L and a 30/60/90 review rhythm.
- 06
Scale
Put capital and people behind what works. Stop what does not.
Before a company exists
- Problem and target user, in one sentence.
- Revenue mechanism — and how it becomes recurring.
- Brand and domain ownership confirmed.
- A written source of truth for the project.
- Repository, deployment and analytics named to standard.
- Data sensitivity classified.
- A phase-zero site with real content and a waitlist.
- MVP acceptance criteria.
- 30 / 60 / 90-day KPIs.
- A continue, stop or scale decision.
Shared across companies
- Domain, DNS, SSL and edge security standards
- Code, review and deployment standards
- An AI gateway for model routing, budgets, logging and fallbacks
- Error monitoring and product analytics tooling
- Brand-neutral UI primitives and security headers
- Access, secrets and cost-review routines
Kept separate — always
- Each product’s user database
- Customer and store data
- Prompts and private knowledge bases
- Business rules and brand voice
- API keys, budgets and logs
- Marketing audiences — no cross-product user graph
Related notes
AI operations ·
AI explains, verified data decides
How we use language models in products where a wrong answer has real consequences — stock levels, visa rules, student access.
Privacy & security ·
Shared infrastructure, separate data
How a group of companies can share technology without building a cross-product user graph.
AI operations ·
Cost discipline for AI products
Per-project budgets, model routing and caching: how we keep AI costs visible across a portfolio.
Work with Oryvelon