The operating system for
portfolio decisions.
A single simulation engine where portfolio intent becomes governed assumptions, assumptions roll forward into consequences, and actual performance explains the variance to drive the next decision.
01 · Plan
Intent becomes a governed assumption.
A growth plan isn't just a volume target; it’s an acquisition cohort with an expected credit policy, pricing tier, and behavioral signature. In Cohort Brain, portfolio intent is structured as explicit, governed assumptions.
- Define behavioral curves (attrition, draw, paydown, default) by segment.
- Set acquisition plans explicitly tied to those behavioral structures.
- Lock and version the assumption set so every downstream result can be traced back.
Machine learning finds the signal. Cohort Brain governs the decision.
Some assumptions are better learned than guessed. Cohort Brain uses trained machine-learning models and statistical calibration to learn from your portfolio's history and derive the inputs others leave to judgment — the shape of a default or paydown curve, how deposit rates move with the policy rate.
But a model never silently steers the forecast. Each model-derived output enters like any other assumption: a named, versioned, governed input you can inspect, challenge and reproduce. The consequential economics — growth, margin, risk and funding — run in the deterministic simulation, not inside an opaque model. The signal informs the decision; the governed economics still own it.
The engine contains no model — only governed inputs.
02 · Forecast
Assumptions roll forward into consequence.
The engine takes those governed assumptions and projects the portfolio forward in time. It preserves the structural reality of the book — cohorts age, balances run off, installments mature, and deposits churn according to the explicit behavioral curves.
Every dollar of forecasted revenue, loss, and funding cost is tied exactly to the month-on-book dynamic of the cohort that generated it. The P&L is an emergent property of the portfolio's structural behavior, not a top-line growth assumption applied in a spreadsheet.
03 · Scenarios & Stress
Understand the boundaries before committing.
A forecast isn't a single path. Cohort Brain applies macro pathways — rate shocks, unemployment curves, funding cost adjustments — as transparent overlays on top of the behavioral engine.
Instantly compare Base, Upside, and Severe Stress scenarios side-by-side. See how policy changes bend the loss curve and impact net interest margin simultaneously. Because it's one model, stress testing isn't a separate exercise; it's a native property of the decision workflow.
04 · Performance & Decision
Variance explains the next decision.
When actuals arrive, they are compared against the locked forecast cohorts. The platform explains variance through the drivers — did volume miss? Did behavior shift? Did pricing change? The evidence immediately informs the next cycle's assumptions.
FINANCE LENS
Total interest, non-interest revenue, and net interest margin trajectory, defensible to the board.
RISK LENS
Vintage loss curves, delinquency roll-rates, and exposure limits tested under macro stress.
GROWTH LENS
Acquisition plans grounded in downstream unit economics, not just top-of-funnel conversion.
TREASURY LENS
Funding requirements and spread analysis mapped against the forward rate path.
One engine. Five distinct portfolios.
The core mechanics of Plan → Forecast → Performance apply to all lending and deposit products, but the behavioral structures are purpose-built for each.
See your next portfolio decision
before you make it.
A focused working session on your portfolio — cards, BNPL, term loans, mortgages or deposits. No pitch deck. No sequence.