Aura: quantitative intelligence
Forecasts with a defined question, time horizon and measurable outcome.
What Aura contributes
Aura is the forecasting engine within Seer. Its role is to turn financial evidence into structured estimates that agents and portfolio services can use. A forecast might estimate a return over five trading sessions or the probability of a precisely defined event.
This differs from a language model writing a confident market opinion. A quantitative forecast has a target, a method, a time horizon and an outcome against which it can be evaluated.
Evidence comes first
Inputs need the right instrument, timing and coverage. A later revision to an economic release cannot be treated as information an earlier forecast already knew. Missing or contradictory evidence changes what the estimate can support.
Forecast uncertainty
An expected return describes an estimate, not a promised result. A range describes part of the predicted distribution under the model’s assumptions. Even a well-calibrated model can be wrong on an individual event.
Seer evaluates forecasts against simpler baselines and subsequent observations. Evaluation shows how forecasts compare with baselines over the same outcomes and time horizons.
From intelligence to a decision
The same forecast can lead to different decisions for two portfolios. Existing exposure, liquidity, cost and risk limits matter. Aura supplies intelligence; portfolio services evaluate allocations; authorization and execution govern financial actions.
The technical reference defines forecast records, evaluation methods and the interface between Aura and portfolio services.