How an AI Voice Agent Runs a Loan Business's Entire Enquiry Pipeline

A note on this case study

This deployment is live with a real client. At their request, we're describing the business generally — as an Australian lending business — rather than naming them, and we're reporting what the system does and how it has changed the day-to-day rather than publishing specific performance figures, since those haven't been formally tracked yet. As soon as we have their sign-off and hard numbers to share, we'll update this page.

The challenge

Loan enquiries are time-sensitive by nature. A prospective borrower fills out a form or calls, and from that moment they're often comparing several lenders at once — whoever responds first, clearly, and credibly tends to win the conversation. For this client, that created a familiar set of pressures:

• Enquiries don't keep business hours. People research and apply for loans in the evening and on weekends, when no one was available to answer.
• Every enquiry needed the same qualifying questions asked consistently — loan purpose, amount, timeframe, basic eligibility — before it was worth a consultant's time.
• Follow-up was manual and easy to drop. A promising enquiry that wasn't chased within a day or two would quietly go cold.
• Pipeline visibility relied on someone remembering to update it — where each enquiry actually was in the process wasn't always current.

None of this is unique to this client — it's the standard shape of the problem for any consultation-driven, enquiry-heavy business. What made it worth solving with an AI voice agent specifically was the combination: high enquiry volume, a fairly repeatable qualifying conversation, and a real cost every time an enquiry sat unanswered.

The solution: Flyvoice on the front end

We deployed Flyvoice, our AI voice agent, as the client's front line for every incoming loan enquiry. Rather than a scripted IVR menu or a simple message-taking service, Flyvoice holds an actual qualifying conversation — asking the same questions a consultant would, in a natural voice, and adapting to how the caller answers.

The brief was specific: the AI should own the repeatable front-end work end to end, and hand off to the client's team only for what genuinely needs a person — reviewing the qualified enquiry and making the actual lending decision.

How the pipeline works in practice

1. The enquiry comes in

Whether it's a phone call or a web enquiry, Flyvoice picks it up immediately — including outside business hours, which is when a meaningful share of loan enquiries arrive.

2. It qualifies the enquiry

Flyvoice asks the client's standard qualifying questions — the kind a consultant would ask first — so that by the time a human is involved, the groundwork is already done.

3. It books the consultation

Qualified enquiries are booked directly into the calendar, with no back-and-forth email or phone tag to arrange a time.

4. It sends follow-up reminders

Automated reminders go out ahead of the booked consultation, and to enquiries that haven't yet converted to a booking — the follow-up that used to depend on someone remembering to do it.

5. It manages the pipeline

Every enquiry's status is kept current automatically as it moves through the stages, so the team can see where things stand without chasing an update.

6. A person takes over for the decision

Once an enquiry is qualified and booked, it's in the hands of the lending business's own team for the part that requires their expertise and judgement — the actual assessment and decision.

What's changed since going live

We're deliberately not publishing invented percentages or dollar figures here — we don't have client-confirmed numbers yet, and we'd rather describe accurately what's changed than dress it up. Qualitatively, the client's team reports:

• Every enquiry gets answered, including the ones that used to arrive after hours and previously waited until the next business day.
• Qualifying is consistent — the same questions, asked the same way, every time, rather than varying by whoever picked up the call.
• Follow-up happens reliably instead of depending on someone remembering to chase an enquiry.
• Less manual admin around booking and pipeline updates, since both now happen automatically as part of the same conversation that qualifies the enquiry.

We'll add concrete before/after metrics to this page once they're tracked and the client has confirmed we can share them.

Why lending is a strong fit for this

Lending and loan-brokerage businesses tend to have exactly the ingredients that make an AI voice agent pay off: enquiries that are time-sensitive, a qualifying process that's repeatable rather than freeform, and a real cost — a lost applicant — every time an enquiry isn't handled quickly. It's a similar dynamic to what we see in trades and field services, where every missed call is a missed job, just with a longer, more structured pipeline behind the first conversation.

If your business runs on inbound enquiries that need fast, consistent handling before a human decision point, this is the pattern to look at. Read the broader picture in AI Agents for Small Business: The Complete Guide, or see the cost of not answering fast enough in What Missed Calls Really Cost Your Business.

Frequently Asked Questions

What does Flyvoice do for a loan business?

Flyvoice answers incoming loan enquiries, asks qualifying questions to establish fit, books consultations into the loan team's calendar, sends automated follow-up reminders, and updates the pipeline as each enquiry moves through its stages — end to end, without manual data entry.

Does the AI handle the entire loan enquiry pipeline on its own?

Flyvoice handles the repeatable stages of the pipeline autonomously — answering, qualifying, booking, and following up — while the lending business's team makes every actual lending decision. The AI manages the process; humans make the judgement calls.

Is a lending business a good fit for an AI voice agent?

Yes. Loan enquiries are time-sensitive, often arrive outside business hours, and involve a fairly consistent set of qualifying questions — exactly the profile where a voice agent adds the most value: fast, consistent answers plus reliable follow-through on every enquiry.

How is this different from a call centre or answering service?

A traditional answering service takes a message. Flyvoice actively qualifies the enquiry against real criteria, books the appointment directly into the calendar, and keeps managing that enquiry through reminders and pipeline updates — work a message-taking service was never built to do.