ADD-ON FEATURE |
Get ready for a new level of Support sophistication with Voice AI. This page contains general information on how the Gladly team will help you get started with Voice AI and help you prepare Gladly for success.
The Gladly Team will help provide a more formal implementation plan during the pre-implementation stages of Voice AI.
Initial requirements
You'll need a Gladly IVR where Gladly can interact with your Customers when they call. Then:
Determine the phone number to be used with Voice AI and confirm which phone number you want Gladly available in.
Determine which IVR to implement Voice AI. Based on the phone number selected for Gladly, determine which IVR Gladly will be activated.
Configure Entry Points and IVR setup
Determine which systems Gladly will need access to: OMS or MDM.
Ensure you have enough Public Answers to answer questions Customer may ask.
Configure Guides to teach Gladly how to behave and respond to Customers. Decide which Guides Gladly should use for which phone numbers, and tell your Gladly Voice implementation contact.
Entry Points and IVR setup
To start, create a layout where Gladly engages with a Customer in the IVR experience. You can insert multiple Gladly experiences in the IVR based on the Guides configured.
Entrypoint | IVR | Where in the IVR | Next IVR node after Gladly AI | Guide to Use |
|---|---|---|---|---|
e.g. 555-555-5555 | e.g. Retale-US-Main | e.g., after the “Main menu” node, if the customer presses “3” for order status | e.g., Handoff to “Main_Agent_Queue” | WISMO (Where is My Order) |
You’ll be asked to confirm the following:
Will you test on the production Entry Point directly with real Customer traffic?
If not, confirm that you allow Gladly AI to set up a new test Entry Point. To allow testing, we will set up Gladly AI to that endpoint.
If you have a UAT/test environment, confirm whether Gladly AI can set up the test Entry Point in your UAT/test environment.
The Entry Point > Voice Rules are set up as intended, and handoffs from Gladly AI will come from the number pointing to the IVR.
Public Answers setup
Ensure you have Public Answers documented for your most frequently asked questions. If you are already using Gladly AI, then it's likely you already have these in place.
If Voice AI is your entry into Gladly capabilities
If you're not currently using other Gladly AI capabilities at the time you implement Voice AI, and if you already have some Public Answers, an analysis of your existing Public Answers must be completed to ensure the best and most useful Public Answers are available for Gladly based on historical Conversations, Topics applied, and other data points.
Gladly Support can help analyze and suggest Public Answers to be created.
Implementation timeline
Below provides the recommended timeline and expected outcome for Voice AI.
Activity | Timeline | Expected Outcome |
|---|---|---|
IVR and Guide setup and internal testing | Week 1 | Voice AI is ready for use |
Limited production test (3-4 hours in production) | Week 1 | Aim for ~2-10% resolution rate |
Production test analysis and Guide optimization | Week 1–2 | Voice AI is optimized for launch |
Provide Agents with a heads-up on upcoming changes | Week 2 | Agents are ready |
Go-live | Week 2 | Aim for ~10–12% resolution rate |
Guide tweaks and next steps planning | Week 2–3 | Aim for ~12–20% resolution |
Implementation Checklist
Configuration - 15 minutes
Voice AI Analysis - 3-4 hours
Guide Tweaks - 60 minutes
Go-live - 30 minutes
Performance Optimization - 30 minutes/week
Continue reviewing Gladly AI results, modify and add Guides and actions to optimize performance weekly.
Commitment expectations
Success is a shared commitment! In a nutshell, here are the expectations from you and Gladly for a successful Voice AI implementation.
.svg?sv=2022-11-02&spr=https&st=2026-03-03T11%3A34%3A34Z&se=2026-03-03T11%3A47%3A34Z&sr=c&sp=r&sig=h0ClSuqgx%2Bydn1ek6hW0Z45K9x0mFONMCINabIgOyn8%3D)
on the top-left corner, then click Contact Gladly Support. Gladly Support's primary operating hours are Monday through Friday, 6 AM to 6 PM Pacific Time, but you can submit a request anytime.