Skip to main content
Book a demoStart free

DRYVN COPILOT

Understand the organization. Prepare the next move. Keep people in control.

Dryvn Copilot helps authorized Change, OD, HR, transformation, and strategy teams analyse organizational context, turn evidence into action, and operate Dryvn through governed workflows.

See Copilot in ProBook a demo
PERMISSION BOUNDARYTransformation Lead4.2–5.0 credits
REQUEST

Prepare a readiness brief for this change.

AUTHORIZED CONTEXT
  • Change impacts
  • Aggregated readiness
  • Organizational evidence
PROPOSED ACTIONManager-readiness intervention draftHuman approval required
A conceptual Dryvn workspace connecting an authorized request to organizational context, evidence, a proposed action, human review, permissions, and usage evidence.

GOVERNED INTELLIGENCE

Not a chatbot beside your HCM. A governed intelligence layer inside Dryvn.

Copilot works within Dryvn's connected organizational model and the requesting user's existing authority. It combines structured context, bounded skills, evidence, and controlled actions without becoming an unrestricted interface over workforce records.

  1. 01Connected organizational model
  2. 02The requesting user's permissions
  3. 03Structured Change and OD evidence
  4. 04Approved Copilot skills
  5. 05Controlled Dryvn APIs
  6. 06Human review
  7. 07Audit and usage records

The intelligence is useful because the context, permissions, evidence, and action boundary stay connected.

COPILOT AT WORK

Turn organizational evidence into a clearer next move.

Copilot applies a different governed skill to each question while preserving the same organizational context and permission boundary.

Change and transformation

  • Review organizational impacts
  • Segment stakeholders
  • Identify readiness gaps
  • Summarize resistance and concerns
  • Draft targeted communications
  • Recommend interventions
  • Prepare steering and programme briefs

Organizational development and effectiveness

  • Synthesize engagement and culture findings
  • Review leadership and manager-effectiveness signals
  • Identify team-health themes
  • Recommend interventions for review
  • Create action and follow-up plans
  • Compare conditions before and after intervention

Organizational insight

  • Explain spans and layers
  • Review vacancies and structural patterns
  • Summarize cross-module evidence
  • Prepare decision briefs
  • Trace findings to supporting evidence
  • Highlight missing context without asserting unsupported causation

PRODUCT WORKSPACE

A workspace built around evidence—not a blank chat window.

REQUESTPrepare a readiness brief for the operating-model change.Authorized as Transformation Lead · read Change and aggregated Effectiveness evidence
AUTHORIZED CONTEXT
  • Programme Atlas
  • 5 affected units
  • 42 positions
  • 3 stakeholder groups
EVIDENCE
  • Impact register · 18 records
  • Readiness pulse · aggregated
  • Communication plan · 7 items
COPILOT FINDINGS
  1. Manager readiness is uneven across two affected units.
  2. Three high-impact roles have no confirmed intervention owner.
  3. The evidence supports a targeted manager briefing before the next readiness check.

Proposed action: create a manager-readiness intervention draft with evidence references.

Human approval required before any Dryvn actionEstimated 4.2–5.0 credits
Conceptual Dryvn Copilot workspace showing permission-bound context and human approval

CONTROLLED ARCHITECTURE

The model can propose. Dryvn decides what is allowed.

Copilot inference sits behind tenant, permission, context, validation, approval, and audit controls. The model receives task-specific context and cannot create its own authority.

  1. 01Authorized Dryvn user
  2. 02Copilot Gateway
  3. 03Tenant and permission policy
  4. 04Context Builder
  5. 05Aggregation, redaction, and pseudonymization
  6. 06Regional managed AI inference
  7. 07Structured result or tool proposal
  8. 08Schema, authorization, and business-rule validation
  9. 09Preview and human approval
  10. 10Dryvn API execution
  11. 11Audit and usage ledger
Every consequential action returns to normal Dryvn authorization, business rules, and audit.

NAMED ACCESS + SHARED CAPACITY

AI capacity your organization can see and control.

Pro includes named access for five authorized people and one organization-wide monthly pool. Administrators decide who can use Copilot, see consumption by user and skill, and control additional capacity.

Named Copilot Users5

Individual authorized accounts. Access follows roles and permissions, not department names.

Copilot Credits1,000 / month

Pooled AI-processing capacity shared by the organization and delivered monthly on monthly and annual plans.

500 additional Copilot Credits — $29/month
  • Usage notifications at 70%, 90%, and 100%
  • No automatic overage charges
  • Copilot pauses new runs at the limit while history remains visible
  • Credits reset monthly and do not roll over
  • Billing administrators control additional purchases
  • Custom capacity and additional named access are available by request
How Copilot Credits are calculated

Copilot Credits = task-specific input tokens ÷ 5,000 + output and reasoning tokens ÷ 1,000

Customer-billable processing

  • The user's request and relevant selected conversation history
  • Authorized tenant context intentionally retrieved for the task
  • Relevant excerpts from customer documents and tool results used by the model
  • Model output and provider-reported reasoning tokens
  • Model processing used to select or propose a Dryvn tool action

Dryvn-absorbed processing

  • Dryvn's fixed system, tool, and skill instructions
  • Dryvn-caused retries, failed provider requests, and duplicate technical calls
  • Internal safety and policy checks
  • Dryvn API calls and retrieval that does not invoke the model

Illustrative examples

  • 2,500 input + 500 output = 1.0 credit
  • 18,000 input + 3,000 output/reasoning = 6.6 credits
  • 100,000 input + 15,000 output/reasoning = 35.0 credits
Compare Pro pricing Read the Copilot terms

PRIVACY + SECURITY

Minimum context. Named access. Human-controlled action.

Copilot reduces unnecessary exposure while preserving enough authorized context to produce useful, evidence-linked work.

  • Permission-bound, tenant-scoped retrieval
  • Minimum-necessary context and aggregation before individual detail
  • Redaction and temporary pseudonymous references
  • No direct database or unrestricted SQL access
  • No long-lived Dryvn API credential given to the model
  • No provider-hosted persistent conversation memory
  • No automatic data-changing action
  • Validation, approval, audit, and usage evidence
  • AI subprocessors disclosed before Customer Data is processed
Copilot security Copilot privacy Subprocessors Copilot terms

RESPONSIBLE USE

Decision support—not automated employment decisions.

Copilot can summarize, draft, analyse, identify gaps, and propose governed actions. Authorized people remain responsible for evidence, judgment, approval, and lawful use.

  • Hiring decisions
  • Termination decisions
  • Promotion or demotion decisions
  • Disciplinary decisions
  • Compensation decisions
  • Medical or psychological assessment
  • Sensitive-trait inference
  • Unlawful profiling or discrimination

Copilot supports the people making organizational decisions. It does not replace them.

ENTERPRISE AI ARCHITECTURE

Customer-controlled boundaries for more complex requirements.

Enterprise customers can establish additional provider, region, identity, retention, budget, and infrastructure boundaries in the applicable Order Form.

Discuss Enterprise Copilot
  • Customer-managed Google Cloud AI project
  • Customer-controlled region, provider configuration, logs, and budget
  • Short-lived workload identity rather than permanent API keys
  • Dedicated or confidential private inference
  • Custom retention and data-boundary requirements
  • Custom Copilot capacity

COPILOT FAQ

Clear answers about access, data, and control.

What is Dryvn Copilot?

Dryvn Copilot is the governed intelligence layer inside Dryvn. It helps authorized users analyse organizational context and evidence, prepare material, and propose controlled Dryvn actions.

Which teams can use it?

Administrators can assign named access to authorized Change, OD, HR, transformation, strategy, programme, or other users. Access follows roles and permissions rather than organizational-unit names.

Does Copilot see every employee record?

No. Dryvn retrieves task-relevant context that the requesting user is authorized to access. Sensitive fields remain subject to existing resource and field permissions, and aggregation is preferred where individual detail is unnecessary.

Can Copilot change Dryvn data?

Copilot can propose an allow-listed action. Dryvn validates the proposal against tenant, user, resource, field, schema, and business rules, shows a preview, and requires explicit human approval before executing a data-changing action through the normal Dryvn API.

Does Copilot make employment decisions?

No. Copilot does not independently make hiring, termination, promotion, discipline, compensation, medical, or other significant employment decisions. Customers remain responsible for human judgment and lawful use.

Are Copilot Credits assigned to individuals?

No. Five named users share one organization-wide monthly pool of 1,000 Copilot Credits on Pro. Administrators control access and can review usage by user and skill.

What happens at the usage limit?

Dryvn notifies administrators at 70%, 90%, and 100%. At 100%, new Copilot runs pause without automatic overage charges; completed outputs and history remain visible until capacity is added or the next monthly allowance begins.

Can we add credits or users?

An authorized billing administrator can add a recurring 500-credit pack for $29/month. Larger or temporary capacity and additional named-user access are available by request.

Is customer data used to train AI models?

Customer data is not used to train or fine-tune shared models. Dryvn also does not train shared Dryvn models on customer prompts, context, or outputs.

Which AI provider does Dryvn use?

Dryvn uses an approved managed Google Cloud AI service for standard Copilot inference. The current provider and processing role are disclosed in the Subprocessors directory.

Where is Copilot data processed?

Standard requests use the configured Google Cloud regional or jurisdictional endpoint. Customer-specific regional architecture can be established where agreed; the exact boundary is described in the applicable service configuration or Order Form.

Can Enterprise customers use a customer-managed AI environment?

Yes. Enterprise arrangements can include a customer-managed Cloud AI project, customer-controlled region and provider configuration, short-lived workload identity, dedicated or confidential private inference, and custom retention or data-boundary requirements.

DRYVN COPILOT

Use organizational context to prepare the next move—with people still in control.

See the full Copilot entitlement in Pro or talk through the security, regional, and capacity requirements for your organization.

See Copilot in ProBook a demo