Deployment
Deploy the AI that fits your organization
Haryts runs its own AI model — no external AI provider receives your documents or questions. Choose the deployment that matches your privacy, residency, and policy requirements.
Deployment postures
Three ways to run — one platform
The same grounded, cited answers, on the AI posture your organization is comfortable with.
Haryts-managed
Haryts answers with its own AI model (Haryts AI), run on hardware Haryts operates. Your questions and the retrieved evidence travel over an encrypted connection to that hardware — no external AI provider receives your documents or questions.
Local AI mode
For strict privacy or residency constraints, Haryts runs on a model inside your environment — both search indexing (embeddings) and generation stay local, so no document content or questions are sent to any external AI provider. Deployed under supervision inside a scoped pilot; answer quality is benchmarked per deployment against the chosen local model.
Evidence-only mode
When policy doesn't permit generative AI for a data class, Haryts still returns cited sources from your documents — no generation required.
Full offline deployment for air-gapped environments is on the roadmap; local AI mode is the current private option and is deployed inside a scoped pilot.
Why it matters
Your knowledge is never hostage to one AI vendor
Because Haryts runs its own model, your documents and questions never go to a third-party AI provider. Your workspace, your documents, and your citations stay intact wherever Haryts runs.
- Your documents and questions never leave to an external AI provider.
- Keep sensitive data classes on evidence-only answers, with no generation at all.
- Run Haryts-managed, or entirely inside your own environment.
- No dependence on a third-party AI vendor for a system your teams rely on.
Deployment capabilities
What's shipping, in pilot, and on the roadmap
The honest state of each deployment capability — with its limitations.
No-LLM Evidence Mode
Still get cited sources when AI generation is unavailable — or when policy doesn't permit generative AI for a data class.
Haryts's own AI model
Your knowledge never passes through a third-party AI vendor, and the platform is not hostage to one.
Local AI mode
For strict privacy or residency constraints, AI generation never leaves your walls.
- Deployed under supervision inside a scoped pilot
- Answer quality depends on the chosen local model; benchmarked per deployment
Full offline deployment package
A complete offline deployment package for air-gapped environments.
- Not verified end-to-end; local AI mode (pilot-ready) is the current private option
Find the deployment posture that fits
We'll map your privacy and residency requirements to a deployment posture during a scoped pilot.