Why we are building in the open, starting from zero

Most startup websites are written for a company that does not exist yet — the one the founders hope to become. The screenshots are mockups, the "customers" are conversations, and the tense is quietly wrong everywhere.

We decided not to do that, for three reasons.

1. Deep tech is a trust business

Nobody buys an autonomous aircraft from a company they do not trust, and trust compounds slowly. If our public claims are accurate when we are small and unimportant, they will be believed when we are large and scrutinised. The cheapest time to establish a reputation for precision is now, when the record is short and easy to audit.

So this site labels things honestly. Research threads are marked active, exploratory, or planned. The roadmap phases beyond the current one say planned, not "coming soon." The gallery has placeholders that say media is pending, because publishing stock imagery of drones we did not build would be exactly the habit we refuse to start.

2. The work is the marketing

An R&D-stage company has no case studies. What it can have is a public engineering log: notes on multi-agent coordination, flight-test write-ups, honest post-mortems. That material does three jobs at once — it sharpens our own thinking, it gives collaborators and professors something real to respond to, and it shows investors how we work rather than telling them.

A hundred pages of adjectives are worth less than one flight log.

3. Honesty is a filter

The people we want — future teammates, research collaborators, early partners — are engineers. Engineers can smell inflation from a distance, and the good ones are repelled by it. A site that says "we are two years from a product, here is the plan, here is the evidence so far" filters for the people who find that exciting.

What this looks like in practice

Concretely, our policy for this site is:

• Every claim maps to an artifact. A prototype, a repository, a test, a written note — or it is labelled a plan.
• Dates are real. News items are dated when they happened. Roadmap dates are targets and say so.
• Status labels are maintained. When a research thread stalls, its label changes. When a milestone slips, the roadmap updates.

None of this is modesty. Our ambitions are large — full-stack autonomy, fleets of coordinated machines, a company built to last decades. We just think the way you earn the right to claims that big is by being ruthlessly accurate about the small ones.