Model risk
Your model risk program decides how models are inventoried, validated and monitored. Exalt Reserve gives it facts to work from: which open models run, where they run, and who signs off.
What the system gives your program.
Models
Open model families such as Llama, Mistral, Qwen, Gemma and DeepSeek, self-hosted in your own cloud account. Which models run is decided with you for each use case.
Where they run
In your account, not on an Exalt-hosted shared service. Your models stay in your environment.
Sources
Every answer shows the source it came from.
Approvals and corrections
The people you name in Exalt Ledger approve material, and correct and approve answers.
A person signs off
Agents are configured around your work, and a person you choose signs off.
Record
Questions, answers, the sources shown and admin actions are recorded with who and when, in your own cloud account. A role you choose exports the log.
Where the line is.
- Validation and sign-off stay with your program. Exalt supports it; it does not run it.
- If a use case also uses an outside AI service, that service's models belong in your inventory too.
Put local AI to work
where it matters most.
Bring your own review questions to the consultation. We answer them against this page.