01
IDEA
Problem hypothesis, market sketch, kill-or-keep call within weeks.
The studio is the platform. The ventures are the products.
i² runs as a small operating studio: one TypeScript stack, one CI pipeline, one set of conventions -every venture inherits the lot. The platform isn't glamorous, but it's the thing that lets a small team ship like a large one.
Venture lifecycle
Four stages, mandatory kill criteria.
Every venture moves through the same lifecycle, with a defined exit criterion at each gate. Sunk-cost rescues are not allowed.
02
PROPOSAL
Validate demand, willingness-to-pay and regulatory feasibility. Mandatory kill criteria are set here.
03
PROJECT
Shared product, engineering, design, legal and ops construct the MVP. Design partners onboarded.
04
PRODUCT
Venture spins out as an independent legal entity, takes follow-on capital and scales -or hits its kill criteria and sunsets cleanly.
Operating principles · People
How we treat each other.
People principles
| Principle | Looks like | Does not look like |
|---|---|---|
| Direct, kind feedback | Say the hard thing early, with care; in person where possible | Sugar-coating, back-channelling, public dunking |
| Default to writing | Decisions and context captured in Notion, async-first | Critical context trapped in DMs or one person's head |
| Ownership | You see it, you fix it or flag it; you finish what you start | "Not my ticket"; throwing work over the wall |
| Curiosity over ego | Change your mind when the data does; ask before you assume | Defending a position past the evidence |
| Sustainable pace | Deep work in long blocks; real recovery; predictable hours | Heroics as a baseline; weekend "saves" that mask broken systems |
| Trust by default | Assume good intent; raise concerns directly with the person | Surveillance, gotcha culture, escalating before talking |
Operating principles · Tech
How we build.
Tech principles
| Principle | Looks like | Does not look like |
|---|---|---|
| Surgical execution | PRs touch one concern; smallest change that solves the problem | Drive-by refactors; speculative scope; "while I'm here…" |
| Test the behaviour | Tests assert outcomes a user can observe | Tests that lock in implementation details |
| Schema-first | Typed contracts defined before code; validated at every boundary | "We'll figure out the shape later"; untyped JSON crossing services |
| Simplicity first | Smallest dependency that survives 3 years; boring tech that ships | Clever abstractions chasing imagined future needs |
| Immutable foundations | Pure functions, idempotent jobs, append-only logs | Hidden mutation; in-place edits to shared state |
| Learn in public | Wins, kills and post-mortems written up for the next venture | Tribal knowledge; lessons relearned each cohort |
Operating principles · Business
How we run the studio.
Business principles
| Principle | Looks like | Does not look like |
|---|---|---|
| Validate before we build | Willingness-to-pay tested before production code | Building "because the idea is exciting" |
| Kill criteria are mandatory | Every venture sets kill conditions at PROPOSAL and honours them | Sunk-cost rescues; "one more quarter" without new evidence |
| Equity, not fees | Alignment with venture outcomes; long-horizon bets | Consulting-hour mindset; optimising for billable activity |
| Compounding playbook | Patterns and platform reused across ventures; each cohort costs less to launch | Greenfielding the same plumbing every time |
| Capital efficiency over scale | Runway is oxygen; every spend earns scrutiny | Hiring or burning ahead of validated demand |
| Portfolio over hero venture | Multiple validated bets in flight at all times | One venture defining the studio's survival |
| Customer trust is the franchise | Honest comms; fix problems before they're noticed; refund without haggling when we're wrong | Hiding outages; deceptive UX; legalistic responses to obvious mistakes |
Engineering defaults
Boring tech that ships.
Defaults, not mandates. Ventures may pick alternatives where the trade-off is justified and documented.
Default stack
TypeScript, React, Node, Postgres, Redis in a monorepo with shared packages. Go for high-throughput backends, Python for ML, Swift / Kotlin for native mobile when the trade-off justifies it.
Schema-first contracts
Typed contracts at every boundary -API, events, forms. Validated at the seam, not in the middle.
CI is sacred
A red main branch is a P0 incident. Tests describe behaviour; we delete tests that only mirror implementation.
Observability before scale
Logs, metrics and traces wired in from day one -not added in panic later.
Accessibility as a baseline
WCAG 2.2 AA target across every venture. Not a backlog item.
Mobile-first
Unless the venture demands otherwise. Most of our users meet us on a phone.
Studio structure
Lean by design.
We hire ahead of pain, not ahead of plan.
Roles at i²
| Role | Scope |
|---|---|
| Founder & CEO | Vision, capital, portfolio strategy |
| Product leadership | Discovery, roadmap, design across ventures |
| Platform engineering | Architecture, shared services, developer experience |
| Venture leads | End-to-end ownership of a single venture |
| Operations & legal | Entity setup, compliance, finance, HR |
| Advisors | Domain-specific, time-boxed, equity-aligned |