> ## Documentation Index
> Fetch the complete documentation index at: https://api.agentflow-enterprise.com/llms.txt
> Use this file to discover all available pages before exploring further.
# Deploy AgentFlow for Multiple Clients as an Agency
> Deploy AgentFlow as your agency RevOps foundation—configure AI-powered lead qualification and CRM handoff pipelines for each client you serve.
As an agency, you can use AgentFlow Enterprise as the shared RevOps foundation beneath your client delivery work. Rather than building a custom qualification stack for every engagement, you configure AgentFlow once per client—branding, workflow rules, and handoff destinations—and deploy a repeatable pipeline that captures leads, scores them with AI assistance, routes them through operator review, and hands them off to whichever CRM or downstream system the client already uses. This guide walks you through what to prepare, how to configure each client deployment, and what to verify before you go live.
## What You Can Deliver to Clients
AgentFlow gives your agency the building blocks to offer structured RevOps as a service. Each client deployment can include:
AI-assisted scoring that evaluates inbound interest against your client's qualification criteria before a human ever touches the record.
A consistent capture layer that replaces disconnected forms, inboxes, and spreadsheets with a single, auditable intake path.
Prioritized operator queues so your client's team focuses first on the leads most likely to convert, not just the most recent ones.
These capabilities reflect AgentFlow's current product direction. Verify live provider behavior—including OpenAI qualification quality and payment readiness—during your pre-deployment checklist before presenting any capability to a client as guaranteed.
## Pre-Deployment Checklist
Work through each step below before you schedule a client launch. Skipping steps here is the most common cause of last-minute delays.
AgentFlow is positioned for multi-client use, but the exact isolation model—data boundaries, access controls, and operational assumptions—must be verified during a private technical review. Confirm that each client's data and workflow configuration cannot bleed into another tenant's environment before you deploy client number two.
Each deployment depends on three external providers. Set them up and confirm they are operational before you touch client-specific settings:
| Provider | Purpose | What to Verify |
| -------------------- | ------------------------------ | ------------------------------------------------------------------------------------- |
| **OpenAI** | AI-assisted lead qualification | Model behavior, response quality, per-request cost, and data handling terms |
| **Stripe or PayPal** | Checkout and billing readiness | Live-mode activation, payment flow across expected states, and webhook event handling |
Do not claim payment functionality is live until you have completed Stripe or PayPal live-mode verification and confirmed payment flows across all expected states. Testing in sandbox mode is not sufficient evidence for a client launch.
Each client deployment needs its own identity and logic. Configure the following before moving to testing:
* **Branding**: Client name, logo, and color scheme in all operator-facing surfaces
* **Qualification criteria**: The rules that determine what constitutes a qualified lead for this client's business context
* **Notification paths**: Where alerts go when a lead needs operator attention
* **Review steps**: Any manual checkpoints your client's team requires before handoff
Confirm that the intake channel—whether a form, landing page, or API endpoint—is correctly wired to AgentFlow's qualification layer. Then confirm the handoff destination: the CRM, operations tool, or workflow system the client uses downstream. A broken handoff discovered after launch reflects on your agency, not the platform.
Submit at least one test lead that mirrors a real inbound signal for this client. Walk it through every stage: intake → AI qualification → operator review → billing readiness check → handoff. Verify that the record arrives correctly in the downstream system and that all notifications fire as expected.
Use a lead scenario that exercises your qualification boundary conditions—a lead that should be rejected, one that should be accepted, and one that falls in the grey zone requiring manual review. Edge cases surface configuration gaps before clients do.
## What Each Client Deployment Needs Configured
Every client is a distinct configuration unit. Use the table below as a per-client setup reference:
| Configuration Area | What to Define | Notes |
| -------------------------- | ------------------------------------------------------- | ------------------------------------------------------------ |
| **Branding** | Logo, name, color scheme, email sender identity | Confirm all operator-facing surfaces reflect client identity |
| **Qualification Criteria** | Scoring rules, fit signals, disqualification conditions | Align with client's sales team before deployment |
| **Billing Setup** | Payment provider connection, plan or pricing structure | Verify live-mode behavior before client billing goes active |
| **CRM Handoff** | Destination system, field mapping, trigger conditions | Test with a real record, not just a schema review |
| **Notification Paths** | Operator alert destinations, escalation rules | Confirm contact details are current at launch |
| **Review Steps** | Manual checkpoints and operator queue configuration | Match the client's existing approval culture |
AgentFlow does not include provider credentials or client-specific configuration out of the box. Each client deployment requires you to supply your own provider accounts (OpenAI, Stripe or PayPal), client branding assets, workflow rules, and CRM handoff details. Plan for this setup time in your project scoping, and verify all providers are operational before presenting any capability to a client as live.
## White-Label and Licensing
White-labeling AgentFlow for client delivery—presenting it as your agency's own product—is a commercial arrangement, not something granted automatically by downloading or deploying the platform. Branding rights, resale rights, client deployment rights, and support obligations must all be agreed in writing before you represent AgentFlow as a white-labeled product to any client.
Do not present AgentFlow as your agency's proprietary platform to clients until you have a written licensing agreement that explicitly grants those rights. Verbal understanding is not sufficient.
If you are evaluating AgentFlow for a white-label program, initiate a private commercial discussion before your first client pitch.
## Realistic Deployment Timelines
Deployment speed depends on how much configuration and verification your client engagement requires.
A single-client pilot with a defined use case, one intake channel, and a straightforward handoff destination can typically be scoped and launched faster than a full rollout. Use a pilot to verify live provider behavior and qualification quality before committing to a broader deployment.
A complete rollout—covering all intake channels, full branding, billing activation, CRM integration, and team training—requires thorough pre-deployment verification across every provider and configuration area. Do not compress this timeline to meet a sales deadline.
Commercial deployment for any client should follow live provider verification. The pre-deployment checklist above is your minimum bar, not an optional step.
## Related Guides
Once your agency deployment is configured, these guides cover the detail behind two of the most important pieces:
* **[AI Qualification](/guides/ai-qualification)** — Understand how AgentFlow's OpenAI-powered scoring works, what it evaluates, and how to tune qualification rules for a specific client context.
* **[RevOps Workflows](/guides/revops-workflows)** — Map the full intake-to-handoff pipeline and connect AgentFlow to your client's downstream CRM or operations system.