Call Our AI: 1300 542 635
One team couldn't live without it. The other wanted to shut it down.
5 min read
TL;DR
A national NDIS consultancy deployed an AI receptionist to handle their growing call volume. The customer service team adopted it as their own virtual team member, handling 4,999 calls and taking 6,700 minutes of front-line work off their plates.
The sales team had a different experience. The AI was built for customer service, and the results show exactly why that matters.
AI receptionist performance · customer service team
4,999
Inbound calls
handled
2,328
Callbacks
arranged
6,700
Minutes of calls
handled
Real performance data from live AI operations. All inbound calls, no outbound campaigns.
The problem
This national consultancy works with NDIS providers across Australia. They're growing, and their phone rings constantly.
The challenge is that two very different types of people call the same number. Existing clients need to reach their account manager, get an update on something, or ask a question. New prospects want to learn about the company's services and whether it's the right fit.
The customer service team handles the first type. The sales team handles the second. And when both groups call at the same time, someone waits, someone gets voicemail, and someone tries a competitor instead.
What the AI actually does
Answers and triages every inbound call
The AI picks up every call, identifies what the caller needs, and works out whether they're an existing client or a new inquiry. No hold music, no voicemail, no menu trees.
Routes existing clients to the right person
For customer service calls, the AI identifies who the caller needs to speak with and either transfers them directly or arranges a callback at a time that works. Callers don't have to explain themselves twice.
Handles routine queries and captures details
Common questions about services, processes, and next steps get answered on the spot. When someone needs follow-up, the AI captures their details and logs everything so the team has full context.
Transfers urgent calls to available team members
When a caller needs to speak with someone immediately, the AI connects them with an available team member in real time. No waiting, no callbacks needed.
80% of callers were happy to talk with the AI
The AI discloses up front that it's an AI assistant. Of 4,999 calls, 80% of people were happy to continue the conversation. The other 20% hung up after the disclosure. That's real data on one of the most common objections businesses have about AI on the phones - and 4 out of 5 callers didn't mind at all.
The results
The customer service team loved it. The AI handled front-line triage, identified callers, arranged callbacks with the right team member, and fielded routine questions. Instead of spending their day answering the phone and working out who needed what, the CS team could focus on actually helping clients.
Why it works
If you've ever built something that works perfectly for one group and drives another group up the wall, you'll know exactly what happened next.
The AI was built for customer service. It was trained to verify who was calling, look up their details, identify what they needed, and route them to the right person. For existing clients, that flow made perfect sense.
For a brand-new prospect who'd just found the company online and wanted to learn about their services? They didn't want to be verified. They didn't want to be looked up in a system. They wanted to say "I'm interested in your services" and hear "Great, let's talk about that."
Instead, they got a customer service triage. The AI tried to look them up, asked if they were an existing client, and walked them through a process designed for someone else entirely. Some callers got confused. Some hung up. The sales team started noticing leads weren't coming through.
They wanted to shut it down.
And honestly? They weren't wrong. The AI was doing exactly what it was built to do. It just wasn't built for them.
6,700 minutes of heavy lifting
While the sales team had legitimate frustrations, the customer service team saw the other side of the story. 6,700 minutes of phone calls handled. 2,328 callbacks arranged to the right person. Hundreds of routine queries answered without anyone picking up the phone. For a team already stretched thin, that's capacity they didn't have before.
4,999
Total calls handled
Every inbound call answered, triaged, and routed. No voicemail, no hold music, no abandoned callers.
2,328
Callbacks arranged
Each one routed to the right team member with full context. The caller never had to explain themselves twice.
6,700
Minutes of call time
Over 110 hours of phone conversations the AI handled so the team didn't have to.
80%
Continued after AI disclosure
The AI identifies itself up front. Four out of five callers were happy to continue the conversation.
What we learned
When AI handles the front line of a business, it's not handling one type of call - it's handling all of them. A single phone number gets existing clients, new leads, vendors, wrong numbers, and everything in between. If the AI only knows how to handle one of those, the rest fall through.
Since this project, every system we build accounts for that reality. Different call types get different flows, different tones, and different outcomes - so no team is left wondering why their calls aren't being handled the way they need.
Why it works
AI works when it's built for the right job. Customer service calls follow patterns. Someone needs to reach a specific person. Someone has a question about their account. Someone needs a callback. The AI handles this well because the workflow is clear: identify, route, resolve.
Sales calls are a different animal. A new prospect wants to feel like they've reached the right place. They want speed, confidence, and a direct path to "let's talk about what you need." Putting them through a customer service triage breaks that experience.
The director had one group who wanted to shut it down and another who said they couldn't live without it. Both were right, because they were using the same tool for different jobs. The CS team got a virtual team member that handled thousands of minutes of front-line work. The sales team got a system that wasn't designed for their workflow. Same AI. Very different experiences.
“I've got one group who kind of want to shut it down. I've got another group who's saying they can't live without it.”
Director, National NDIS Consultancy
4,999 calls handled. 2,328 callbacks arranged. 6,700 minutes of conversations the team didn't have to take. The customer service team calls it their virtual team member, and the numbers show why. When AI matches the workflow, the results speak for themselves.
Build AI that fits how your team actually works.
See how AI could handle your front line, backed by data from
171,688+ real conversations.
Earlibird AI
AI systems for revenue growth and operational efficiency.
Call Our AI: 1300 542 635
Copyright 2026. Earlibird - AI Made Simple. All Rights Reserved.