Visible provenance
Supported answers retain the source context a reviewer can inspect.
Trust through evidence and control
Inspect the source. Understand the boundary. Keep people in control.
Orchestri is designed to make recommendations reviewable: show the observed context, retain the relevant sources, surface uncertainty or missing information, and separate recommendation from approval and execution.
Action control
Recommend → Preview → Approve → Execute → Audit
Not every recommendation is executable. Supported actions remain subject to the product workflow, configured permissions, and human approval.
- 01
Recommend
Prepare a response from supported customer and revenue context.
- 02
Preview
Show the proposed work, supporting evidence, and known gaps.
- 03
Approve
Leave the consequential decision with an accountable person.
- 04
Execute
Perform only a supported action within configured permissions.
- 05
Audit
Retain the action context where the product workflow supports it.
Technical proof
A recommendation should be inspectable before it is actionable.
Orchestri's differentiation comes from governed commercial context—not exclusive dependence on one model provider.
Data freshness
Where supported, the experience shows refresh health, timestamps, or stale and missing context rather than implying real-time coverage.
Permission-aware context
Recommendations use the records and workflow context available through configured access.
Human review
Consequential pricing, legal, forecast, contract, migration, and customer-impacting work remains reviewable.
Published Overlay boundary
Read supported CRM records without writing changes back.
Current Salesforce and HubSpot Overlay access is read-only. This boundary prevents Orchestri from changing source CRM records; it does not make the integration risk-free.
Salesforce
Accounts, Contacts, Opportunities
Orchestri can use these supported records for views, answers, risks, and recommendations. The source system remains unchanged.
HubSpot
Companies, Contacts, Deals
Orchestri can use these supported records for views, answers, risks, and recommendations. The source system remains unchanged.
What read-only does—and does not—mean
Read-only lowers the risk of source-record corruption, preserves system ownership, and creates a bounded starting point. Access, authentication, privacy, data quality, synchronization, and governance still require review for the specific environment.
Security-review honesty
Unverified policy details are not published as assurances.
Orchestri does not publish the topics below as assurances until each answer is approved for customer use. Confirm current details directly during a qualified security review.
SOC 2 and assurance status
Not asserted on this public page.
Infrastructure, storage, backup, and AI-processing regions
Not asserted on this public page.
Retention and deletion windows
Not asserted on this public page.
Current subprocessors and their roles
Not asserted on this public page.
AI provider selection, training, and retention terms
Not asserted on this public page.
Security contact and vulnerability reporting
Not asserted on this public page.
This page does not claim SOC 2 certification, a particular data residency model, exact retention or deletion windows, zero-retention AI processing, a named subprocessor list, or a vulnerability response commitment.
Public documents
Review the current website policies.
These documents provide the public legal and privacy language currently available on the site.