Industry solution
Decisions you can explain, on data you can trace.
For teams that must show how a figure was produced and why a model decided as it did.
The problem
What teams are dealing with.
- Reconciliations and regulatory reports are assembled from extracts by hand, under deadline.
- A risk model is only useful if its decisions can be explained to a customer and a supervisor.
- Every new tool is another system to control, test and evidence.
Why it stays unsolved
Existing approaches fragment the workflow.
- Data preparation, modelling, reporting and model risk management live in separate systems with separate records.
- The explanation of a model is documented apart from the model that is running.
The DooHorizon solution
DooHorizon keeps the data path and the model evidence in one environment.
FlowDoo
Status: AvailableAutomates extraction, reconciliation and report preparation with a log for every run.
About FlowDooMLDoo
Status: PreviewBuilds risk models with per-decision explanations and a Model Passport.
About MLDooVizzDoo
Status: Public BetaPresents exposure and performance to management.
About VizzDooGovernanceDoo
Status: Coming SoonIs designed to support model inventory, approval and audit.
About GovernanceDoo
Reference architecture
How the pieces are arranged.
- Core systemsLedgersFiles
- FlowDoo
- ReconciliationMLDoo risk model
- Review
- ReportsDecisions
Example business workflow
An explainable decision with its evidence attached.
- Application data
- Validate
- Score risk
- Explain the score
- Human review
- Record the decision
Business outcomes
What this lets you do.
- Reduce manual assembly in reconciliation and reporting.
- Explain an individual model decision.
- Give model risk reviewers the evidence in one place.
- Keep data inside your own infrastructure with a private deployment, where offered.
Governance and security
How it stays under control.
- DooHorizon does not claim any financial-sector certification or regulatory approval.
- Self-hosting keeps data on infrastructure you control.
- Audit trails and lineage are built into the products available today.
Start with this use case.
Start with one use case. Expand across the ecosystem.