Workflow Map
Use this page when you want one compact view of every major workflow, its input shape, the right first command, and the first successful output you should expect.
Workflow inventory
| Workflow | Business question | Input shape | First command | First win | Next doc |
|---|---|---|---|---|---|
| Vintage Curves | How are losses accumulating by cohort over time? | Vintage long format | cranalytics quickstart -> choose Vintage Curve Fitting |
Fitted curve and ultimate loss estimate | Vintage Analysis Tutorial |
| Lifetime Loss Forecasting | What lifetime loss does this portfolio imply under transition assumptions or loan history? | Portfolio snapshot + transition matrix or loan history ledger | cranalytics quickstart -> choose Lifetime Loss Forecasting |
Reserve estimate on a mock portfolio | Lifetime Loss Forecasting Tutorial |
| FICO Segmentation | How is the book distributed by score band and simple risk grade? | Portfolio snapshot | cranalytics quickstart -> choose FICO Segmentation |
FICO-band table and mix diagnostics | FICO Segmentation Tutorial |
| Feature Analytics | Which variables have enough signal to justify modeling work? | Loan-level modeling frame | cranalytics quickstart -> choose Feature Analytics |
Ranked feature table | Feature Analytics Tutorial |
| ML Modeling | How well does a predictive model backtest over time? | Loan-month panel or labeled modeling frame | cranalytics quickstart -> choose ML Modeling |
Fold-level backtest summary table | ML Modeling Tutorial |
| Rollforward Workflow | Is my aggregated rollforward data ready, and which variant should I trust operationally? | Monthly aggregated rollforward data | cranalytics rollforward-readiness your_rollforward_data.csv --output-dir rollforward_readiness_out |
Readiness report directory and diagnostics | Rollforward Workflow API |
| Survival Analysis | How does default or prepayment timing evolve over time? | Time-to-event dataset | python -m cranalytics.examples.core_survival |
Survival curve / hazard timing view | Survival Analysis Tutorial |
Start from the data you already have
| If your starting data looks like... | Start here |
|---|---|
| Loan-level portfolio snapshot | Lifetime Loss Forecasting or FICO Segmentation |
| Loan history panel by observation date | Lifetime Loss Forecasting or ML Modeling |
| Loan-level booked loan file plus early-performance flag | Feature Analytics |
| Loan-month performance panel | ML Modeling |
| Monthly aggregated flows by segment and MOB | Rollforward Workflow |
| Cohort/vintage loss observations | Vintage Curves |
| Time-to-event dates and outcomes | Survival Analysis |
Documentation freshness checklist
Review these together whenever a new onboarding-facing workflow or command ships:
README.mddocs/index.mddocs/getting_started.mddocs/choose_your_path.mddocs/reference/workflow_map.mddocs/reference/module_map.mddocs/reference/public_api_map.mddocs/reference/cli_reference.md- the relevant tutorial page
- the relevant packaged example under
src/cranalytics/examples/
Treat this checklist as part of the definition of done for onboarding-facing changes. If one of these surfaces changes, the rest should be reviewed in the same pass.