ROOT — the moat
ROOT is the reference data model ("Datamode") of the Amazon — calibrated once, then reused. It is both the defensibility and the cost-reduction engine of the business.
Measure once, infer many
The basin has ~15 ecosystem subtypes (per Peru's MINAM ecosystem map + global classification). CeibaQ builds a calibrated reference for each subtype on a small number of well-sited concessions (~8–10 concessions cover all subtypes via natural terrain gradients). After that, most hectares are identified and read by the model plus a light drone pass — not measured from scratch.
What it cuts
A strong ROOT reduces the field-ops portion of delivery cost by roughly 60–80%, which is about 40–55% off blended cost-to-serve; combined with scale, total cost-to-serve falls from a high pilot figure toward ~$2/ha at scale.
What it costs (planned)
A one-time build of ~$10.7M (sensor network + a SMEAR-class flux supertower + AI/calibration + reference libraries + integration of 20-yr government data) plus ~$1.8M/yr to run. Designed to be funded largely by non-dilutive grant/MDB/results-based capital — so equity funds commercial traction, not the science capex.
Why it's a moat, not just a cost
- Direct measurement (flux + eDNA + LiDAR) that satellite players structurally cannot reproduce.
- 20 years of government data in-kind via institutional agreements — not buyable on the open market.
- Built-once reference libraries (species, eDNA barcodes, acoustic signatures) with zero marginal cost to reuse.
- Amortizes to near-zero per hectare at scale, while remaining a multi-million-dollar, multi-year barrier to entry.
Status: ROOT is to be built. The pre-seed funds a lean proof-of-concept on one subtype on the owned concession (full ROOT + flux supertower come at seed, largely non-dilutive).