The thesis
There's a developer-partner in Colombia's Valle del Cauca who sits on something valuable: land access across a regional network of sugarcane (caña) growers, plus the EPC relationships, utility connections, and power-purchase agreements to actually build on it. The play is to structure one financing vehicle that funds a standardized ~1 MW solar farm, prove it with a pilot, then replicate that exact product across the grower network — solar-farm-as-a-platform, distributed through one partner's relationships. The partner brings the channel; I structure the capital.
What I do here
I'm the operator on the structuring side again — building the fund model, designing the vehicle and its governance, and working the cross-border legal and tax structuring. The most useful thing I did early was an honest diagnosis: the asset is fine, but the deal as first drawn doesn't clear the return hurdle, and the reason is the capital structure, not the sunshine. So I traced it to the financing — tenor and coverage in the ramp — and defined the levers to re-engineer it into something an investor can actually back.
What's built
- ◦An institutional fund model — development-bank-grade, validated two independent ways, with an interactive lever tool for testing restructuring options live instead of arguing about them.
- ◦A cross-border structuring memo — resolving the central tension in deals like this: how to capture Colombia's domestic renewable-energy tax incentive while running an offshore, dollar-denominated vehicle. The answer is a layered holding chain that uses a tax treaty to optimize withholding — real structuring work, not a template.
- ◦A meeting-to-action loop — an AI skill that pulls call recordings into structured notes, extracts the action items, and pushes them into project management, so the relationship never loses momentum between calls.
Where it stands
The earliest stage of anything in the constellation. Principles are aligned and there's an agreed timeline, but no signed contract yet — an NDA and a first model exchanged. A go/no-go decision is targeted around mid-2026, with a launch shortly after if it proceeds. The pilot farm is physically built but waiting on the local utility's interconnection process to go live. There's also an interesting embedded-IP angle: the partner built his own plant-monitoring software that reads inverter data to catch faults and reconcile utility bills, already with paying customers.
Nearby stars
Sister effort to Orchidea — kept as a deliberately separate workstream (different jurisdiction, different counterparties) but sharing the same operator and the same Alpicat tooling.