01 / Overview
What a weather market taught us about organizations.
HEAT began with a concrete problem: turn weather observations into a view of uncertainty, compare that view with a market, and evaluate the result against an external settlement record. Each part has to agree on the instrument, the place, the time, and the evidence.
That operating problem keeps producing organizational questions. Which observation should be trusted? Who can act on it? How much authority does a decision carry? When should the system wait? And how does a failure become a rule the next worker can remember?
HEAT Trading is our applied research setting for answering those questions. This preview follows the path from a weather-derivatives system toward an organization whose beliefs, resources, responsibilities, and results remain explicit over time.
02 / Capabilities
The work behind the research.
Observe with provenance
Connect weather readings, forecasts, and market observations to their sources and time semantics. A current sensor reading and the official record used for settlement serve different purposes.
Decide within a mandate
Treat price, uncertainty, liquidity, instrument identity, and risk limits as part of the decision. Waiting, refusing, and escalating are meaningful outcomes of a controlled process.
Learn from external outcomes
Connect later settlement to the evidence, model state, and decision that came before it. Preserve failures as reviewable lessons, then test whether a change improves the next cycle.
03 / How it works
Intelligence, organized.
- 01
Observe & form a belief
Interpret weather and market evidence in the context of a specific contract.
- 02
Propose & review
Evaluate a possible action against its price, uncertainty, and authority.
- 03
Act, wait, or escalate
Take the permitted next step and retain its decision record.
- 04
Settle, grade & revise
Compare with the external outcome and update the operating knowledge.
The research question
Can an organization learn without losing its memory or its mandate?
We are exploring a durable graph of responsibility: objectives, beliefs, delegated authority, resources, commitments, and outcomes that persist while models and workers change.
- 01Mandate
- 02Evidence
- 03Decision
- 04Accountability
04 / Research frontier
A research preview with a concrete starting point.
The existing HEAT system contains data feeds, orchestration, predictions, risk checks, archived market books, and settlement reconciliation. Its research artifacts also document forecasting and instrument-identity failures, and the work of evaluating repairs.
The broader Autonomous Organization is being formalized. Some judgment, role definitions, exception handling, and changes to authority still depend on the founder. Moving that knowledge into explicit, reviewable organizational state is part of the research.
The question is how to extend the time a system can operate usefully inside a mandate while keeping intervention, review, and accountability possible. This page presents the research direction and the operating lessons behind it.
