Skip to main content

Onboarding & implementation

This page explains how a typical onboarding and implementation process works when a new customer starts using CallPilot.ai.

The goal is to give you a clear picture of:

  • What needs to be prepared.
  • Who does what (your team vs the CallPilot.ai team).
  • The usual phases from kick-off to go-live.

Objectives of the onboarding​

By the end of onboarding, you should have:

  • At least one end-to-end campaign running (voice and/or WhatsApp).
  • Basic dashboards and reports configured.
  • A clear process for testing, go-live, and ongoing improvements.

Roles and responsibilities​

Onboarding usually involves two sides:

Your team​

  • Sponsor / Business owner
    • Defines goals and success metrics (e.g. increase contact rate, reduce manual calls).
  • Operations / Supervisors
    • Help design scripts and flows.
    • Validate that campaigns behave as expected.
  • Admins / IT
    • Provide access to systems (CRMs, core systems, etc.).
    • Approve numbers, channels, and security aspects.
  • Developers / Integrators (if applicable)
    • Connect CallPilot.ai to your systems via APIs and webhooks.

CallPilot.ai team​

  • Guides the onboarding process.
  • Helps design flows and campaigns aligned with your goals.
  • Assists with technical integration and best practices.
  • Supports testing, go-live, and early optimization.

Typical onboarding phases​

Every customer is different, but a common onboarding flow looks like this:

  1. Discovery & goals
  2. Setup & configuration
  3. Integration (optional)
  4. Testing & UAT
  5. Training
  6. Go-live & early optimization

You can adapt this to your internal process.

Phase 1 – Discovery & goals​

In this phase, we clarify what you want to achieve with CallPilot.ai.

Typical questions:

  • Which use cases do you want to start with?
    • Outbound collections? Renewals? NPS? Inbound triage?
  • What are the current volumes (calls/messages per day/week)?
  • What does success look like in the first 30–90 days?
    • Higher contact rate, lower handling time, better data quality, etc.

Deliverables:

  • Prioritized list of initial campaigns (e.g. “Collections”, “Renewals”).
  • Agreement on KPIs to monitor after go-live. TO BE DEFINED CON FELIPE/CAMILO

Phase 2 – Setup & configuration​

Here we prepare your CallPilot.ai environment.

Typical steps:

  • Create or confirm your organization / account.
  • Configure channels:
    • Voice numbers.
    • WhatsApp senders (Note: must be configured with your own Meta account).
  • Define users:
    • Add users to your callpilot site.
  • Configure any basic settings required by your organization.

Deliverables:

  • Your team can sign in and access the correct workspace.
  • At least one voice or WhatsApp channel ready for testing. TO BE DEFINED W/ CyF

Phase 3 – Integration (optional but recommended)​

If you plan to connect CallPilot.ai to other systems (CRM, core system, data warehouse, etc.), this is where we define and implement the integration.

Typical tasks:

  • Decide how contacts will be:
    • Imported (batch/API).
    • Updated with outcomes.
  • Define webhooks:
    • Which events to send (e.g. call completed, payment promised, survey answered).
  • Implement and test API calls on your side.

Related docs:

  • Developer Guide (APIs, webhooks, examples).
  • API Reference (/api).

Phase 4 – Testing & UAT​

Before going live with real customers, we run tests.

Functional testing​

  • Build test workflows that reflect your real scenarios.
  • Use a testing widget for your case scenarios.
  • Validate:
    • Messages and prompts.
    • Branching logic.
    • Handoffs to agents or external systems.

UAT (User Acceptance Testing)​

  • Run limited campaigns with a small subset of real contacts, if appropriate.
  • Collect feedback from:
    • Operations/supervisors.
    • Agents (if involved).
    • Business stakeholders.

Deliverables:

  • Approved flows and campaign configuration ready for production.
  • A clear list of adjustments to be made before go-live.

Phase 5 – Training​

To ensure adoption, different roles may need different training:

  • Knowledge Base

    • Learn how to structure and upload documents so the AI interprets them correctly.
    • Understand best practices to reduce hallucinations and keep answers consistent.
    • Maintain and update the content library for accuracy over time.

  • Workflow

    • Learn how to design effective prompts and conversation paths.
    • Build flows that adapt to diverse customer cases (sales, claims, support, quoting).
    • Optimize logic and routing to improve resolution and reduce agent load.

  • Reporting

    • Learn how to use AI-driven insights to improve campaign and workflow outcomes.
    • Understand performance dashboards, drop-off points, and optimization opportunities.
    • Use analytics to refine messaging, prompts, and business rules.