Katey: Hi and welcome to Broker Tools, where we unpack the tools, systems and strategies that help brokers optimise the way they work. In this episode of How to Use Software, we're catching up with Brenton from Quickli. If you've been wanting to automate your pre-assessment process, this episode is for you. Welcome to the pod, Brenton.
Brenton: Hi, Katey. Thanks for having me — I'm very excited to show you all that Quickli has to offer, especially Pro. If I can do it within the hour or 45 minutes... I always pre-warn everyone on my webinars now, because the team keeps adding more features, so it's hard to keep talks within 45 minutes to an hour. It's a great problem to have, though, because it means brokers get more features. So I'm excited to run you through Quickli Pro and our AI features.
Katey: Nice — and for context, Quickli didn't actually start with Pro, it started with the servicing calculators, is that right?
Brenton: Yes, correct. Quickli came about in 2021, originally to solve the servicing headache. Before Quickli, brokers were workshopping at least three lenders for a client — sometimes four or five — which meant filling out three, four, five different lender spreadsheets by hand. On a complex deal that could take hours, and if you're a new broker trying to get your head around 50-plus lenders' spreadsheets, that's a huge job in itself. With Quickli you put the information in once — policies, procedures, income shading, LVRs, HEMs, everything needed to get to servicing — and export it directly into the lenders' calculators, using their own policies and shading, so you only enter it once instead of three, four or five times. It's evolved a lot over the last five years — there are policy libraries in there now, and we've since launched Quickli Pro, which has a bunch of features including a few AI features, which we'll show off today.
Katey: Nice, and I guess you also have the widget people can put onto their own website?
Brenton: Yes, correct — that's one of a bunch of offerings. We do website widgets for businesses, which can be white-labelled with your company's colours and look and feel. The benefit of using Quickli's widget is that it's connected to our core platform, so if there's any change in the background — say a lender's HEMs — that updates on your widget too, so a client comparing results there stays accurate as long as they've entered their information correctly.
Katey: Nice. I like to think about things from a systems and process point of view — when a client comes in, I've found brokers use Quickli at different stages, but often the best time is on the discovery call: you put in some details and save them, which helps start the process of presenting options while you're actually on the call.
Brenton: Correct — that's still how a lot of brokers use Quickli. They do the initial discovery call, and while they're on the phone — on speaker, headphones in — they're putting in that information, because they get instant results from 50-plus lenders alongside them. That gives a quick, instant idea of whether a deal looks viable, before requesting documents — and nobody wants to request 50 documents from a client if it's not going to service. Now that we have Pro, some brokers are doing that initial discovery call via Jiffi instead, using voice-to-text — you talk to Jiffi and it builds the scenario for you, so there's no manual keying-in required.
Katey: Nice — do you want to walk through it? I think that's the whole purpose here, so people can understand how to use their software, especially Quickli.
Brenton: Yeah, I'll share my screen and give you a demo. I'll demonstrate the AI features today — Jiffi AI, Doc Handler and Loan Submission Notes — but since this podcast is about software and automation, I want to point out you don't just get the AI features on Quickli Pro. There are four other powerful features too: our SMSF tool, our Doc Specialist tool, custom branding, and organisation analytics. And there's an exciting feature coming — the missing puzzle piece I'll give you a sneak peek of — our Fact Find and Doc Collection.
The whole premise with Quickli Pro, especially the AI features, is to shrink the preliminary assessment time down. Think of it this way: at the back end you've got the start — researching policy and product, getting documents from the client, renaming them, writing up loan notes — that whole process can take hours. We think we've already solved the back end, because once you're in Quickli you put information in once and your servicing's done, ready to export to whichever lenders you like. We're now trying to shrink the front end and meet in the middle — we're pretty close. Once Fact Find and Doc Collection is released, we hope the two will meet, and the whole process — right through to an export ready for the lender — could be done within about five minutes, and automated.
There are aggregators with fact finds, or people using external fact-find providers, and there's no right or wrong — but a question I get asked constantly since Pro was released is why use Quickli's own Fact Find and Doc Collection. My answer: you're doing everything else in Quickli anyway — we already research policy and products, do admin tasks, rename documents, extract data, run servicing and write loan notes. So imagine sending this to a client on a Friday afternoon — if they fill it out over the weekend, you come in Monday and all of that's already done automatically, instead of spending half your day sorting and keying things in. You could come in ready to submit to the lender straight away.
So what I'll go through today — there's no right or wrong. I get asked a lot whether you should start with Jiffi or Doc Handler. There isn't a right answer, but I'll show a couple of examples of how brokers I speak to are using these tools.
Katey: Cool — and I think this is the key thing: one of the things we're aware of in the industry is how many pieces of software exist to do one job, and what you're trying to do is reduce that. If you already own Quickli and you're doing the pre-assessment process — which is what Jiffi does at the start — you can naturally flow on into the Fact Find and the rest of the processing.
Brenton: Yes, that's exactly right, and it's one of the biggest pain points I hear. I spoke to a Quickli user last week at National Finance Brokers Day who said they were using eight or nine different pieces of software and it was getting too much — they had Pro but weren't using it to full potential and wanted to cut down. After a quick demo he said he could probably drop three of his other pieces of software just from that first call, because he hadn't realised Pro already covered them. So exactly what you said — we're trying to make things as simple as possible. There are over 14,000 brokers on Quickli now, two-thirds of the industry, so it makes sense to be able to do everything in here — it makes brokers' jobs easier and more efficient.
Katey: Yeah, we're trying to do what we call a reduction of software fatigue.
Brenton: Yes — it's a big problem, and hopefully once people see this we've solved it for them.
For the sake of time, just before this call I put together a workshop scenario, because while Jiffi is there to answer policy or product questions, I also want to show it goes above and beyond that. Think of Jiffi as your credit analyst — you can fully workshop deals with it like a staff member that's there 24/7 and doesn't need a break. It can run funds positions, grab documents, draft emails, check SLAs and postcode restrictions — all the little tasks that otherwise mean logging into broker portals and going here, there and everywhere. Jiffi can research a policy or product question, then go further and do a lot of the actual pre-assessment work to see if a deal will work.
I'll show you quickly — a straightforward policy question, like which lenders allow apartments under 50 square metres, gets a quick, easy-to-read answer within five to ten seconds, the traditional way an analyst would work through 'can I do this, can I not', because we hold all the product data in Quickli — you could even ask for a 25-product comparison. But what I'll really show is how brokers use this in their workflow day to day. The most common question I get is whether to start with Jiffi or Doc Handler — there's no right or wrong. You can still key things in on the discovery call as before, and I've spoken to brokers who use the voice-to-text feature on their headphones, repeating the client's numbers back to them, and Jiffi transcribes and builds the scenario — even pre-filling the servicing — so you skip a step and can see instantly whether it's going to service.
Before I go through the scenario, a bit of context: I'm asking Jiffi to do ten to fifteen different things here, and it won't get overwhelmed. There's a nurse on maternity leave returning to work — is that okay? A second applicant who's a self-employed sole trader, where we want to use the most recent year's net profit. There's an existing owner-occupied property they want to convert to an investment — can they refinance and avoid LMI? I've included their liabilities, dependents, and the purchase price of the new property, again wanting to avoid LMI, plus asked it to check postcode restrictions — in this case I'm in New South Wales looking at a Queensland postcode, so it's worth checking for restrictions. It'll tell me which lenders suit, run servicing (connected to Quickli), and can even grab documents — I've asked for an ANZ discharge and to check SLAs.
This is the point where, if you're on a discovery call without documents yet and want to workshop the deal, I'd recommend starting with Jiffi — it'll create a scenario that flows into Doc Handler and then loan notes. But you could just as easily start with Doc Handler, which I'll show shortly. First, maternity leave: generally fine across the panel. Jiffi goes further than a yes/no — it flags that she'll need two payslips showing current base salary, year-to-date figures and employer ABN, possibly an employer letter, and — because she's returning from leave — it's even asked about childcare costs. That's what I mean by Jiffi being a genuine credit analyst: it thinks of things you might not, which is useful whether you're a new broker still learning or an experienced broker running a big team.
There's the sole trader question about using the latest year's income — it gives an easy-to-read summary of conditions, required documents and caveats, flagging a 60% income variance that some lenders may apply a variance test against. It's worked out the funds position on the existing property — enough equity, with $93,000 surplus cash remaining after the cash-out and refinance to avoid LMI. So you can see how much detail Jiffi goes into if you ask it to workshop a scenario — postcode restrictions checked, the ANZ discharge form retrieved (handy since discharge forms are often hard to get from lenders directly), and servicing run — it's telling me this isn't currently servicing, and crucially it explains why: the DTI is 7 and there isn't quite enough income, flagging the credit cards too, and suggesting what could help rather than just stopping at 'no'.
It's picked up that she mentioned overtime — she's a nurse, and many lenders allow 100% of overtime — prompting a follow-up question you might ask the client to help get the deal over the line. It finishes with SLA turnaround, a summary, and recommended next steps, so you can make an informed call to your client. That's the level of detail Jiffi can go into.
One more thing before moving on — the settings and custom instructions, which I think are underused. Just like using AI outside of broking, you tell it exactly how you want it to answer. If you have a lender you'd rather avoid, you can note that; if you want alt-doc results limited to specific lenders like Pepper, Bluestone or RedZed, you can set that too. Jiffi can also draft emails, so you can set a tone for how you want them written — I could ask it right now to draft an email to an ANZ BDM about a specific scenario, and it would, based on everything already in there. I can't stress it enough — put in as many instructions as possible.
Katey: And if you're familiar with Claude, it's almost like your projects or your co-working space there.
Brenton: Yes, correct — and you can give it memories, which it saves as custom instructions, so I'd encourage everyone to do exactly that.
A few other handy things: you can share a scenario with your team — useful if you're a team of five, say, and want to know who does what. There's no right or wrong; maybe the broker does the first call, then hands it over — 'I've got everything here, even the SLAs and the discharge, and a few lenders that look okay, can you put the deal together?' With Doc Handler, which I'll show shortly, your team member can pull that application together quickly. You can also share the scenario or print it to PDF.
Katey: With the folder at the bottom — are you adding that, or is it produced by Jiffi for you?
Brenton: That's produced by Jiffi. If I click on it now, it's actually created a scenario for you within Quickli. Before Pro, you'd do your first call and key it in as we described, or take the call in the car and key it in later from memory. Using Jiffi like this just creates the scenario directly, so there's one less thing to do.
Katey: Excellent — and can you use Quickli on your mobile phone?
Brenton: It's not the best on mobile at the moment, and we're aware of that — there could be improvements coming, we're working on it. It got pushed down the priority list a bit with the Pro release this year, but it's coming back up now.
Katey: Cool — and sorry, just a side question: if someone uses something like Whisper Flow to record and transcribe conversations, could that be added in?
Brenton: Ah, the voice recording — no, there's no direct connection for that, but if you've got the transcript, you can just copy and paste it straight in, and do exactly what we've just gone through.
Katey: Oh okay, cool — so there's no way to attach files in this chat?
Brenton: No, not in Jiffi's chat at this stage — that's something we're aware of too, we can't plug that in yet. But copying and pasting an external transcript works fine.
Katey: Cool, cool, cool — awesome.
Brenton: So I'll show you now — going back to that nurse-and-electrician scenario. This is based purely on what you've told Jiffi — you could go back in and update the names, say to Adam and Sarah, which is no big deal, but it's a very good starting base. If you've got everything from the client already, you likely won't need to add anything else into Quickli.
We're seeing a lot of red here, which Jiffi flagged as not servicing, but with prompts on how to get it over the line. All the information — dependents, income for the sole trader, the existing properties with auto-calculated LVRs, liabilities — is already keyed in for you. If this were real, I could hide a few fields to get closer to servicing, and because Jiffi flagged some overtime, adding that in might tip a few fields to green — or send you back to the drawing board.
That's Jiffi. Now, Doc Handler — a feature that's had some recent updates. The flow: once Jiffi suggests a couple of lenders will service, you get documents from the client and drag and drop them in, which prefills anything needing an update and analyses income, adding it to that scenario. I'll start a fresh scenario here for clarity, since this is a completely different case. With Doc Handler you can either create a brand-new scenario and upload from there, or skip the scenario and go straight into the toolbar to upload new documents — there's no right or wrong, though if there's no scenario yet I'd recommend going straight in, since it'll ask afterwards whether to push everything into a new one.
Right now, before Fact Find and Doc Collection is released, you still need to get documents from the client however you currently do — but once that's live, collection will be automatic. These documents were deliberately given unhelpful names — you can upload up to 100 at once, which is handy since nobody wants to rename that many by hand. It renames them and extracts income data in two parts — once this finishes loading I'll walk through it, and come back to the personalisation settings after.
Katey: So what Doc Handler is doing is recognising the documents submitted — you gave them names like AAA and JJ, but it's recognised, say, that this one's a passport, correct?
Brenton: Correct — it's renamed it 'Passport'.
Katey: Can you set your own naming conventions?
Brenton: Yes, definitely. Something I think anyone listening will love: this is going to be integrated into the fact find for clients. Most brokers I speak to — myself included, I'm a broker too — find getting documents from a client is always the hardest part, and it can end up in the too-hard basket. You'll be able to tell clients they can just drag files into the fact find, even messy ones like these, even images, and we'll convert and organise them automatically.
You can personalise naming conventions in settings — spacing, date formatting, name formatting, number formatting, and there are around 88 configurable fields, unless the team's added more. Some people tweak as they go, others set aside time to overhaul everything — you might strip it back to just the account number and statement date. If you overdo it, everything saves automatically, so you can just revert to default and start again. You can also set conventions per major bank — like abbreviating National Australia Bank to 'NAB' so statement names don't get too long. On top of that, we convert images to PDF, redact tax file numbers, and let the broker choose how long documents are stored within the scenario.
Katey: Perfect — which is handy, since I understand a driver's licence can only be held for a certain period too.
Brenton: Yes, it's entirely up to the broker — 24 hours or a year. We take privacy and security seriously at Quickli: if you set 24 hours, documents are deleted after that and we can't provide another copy, so I always flag that. If you're still mid-way through collecting documents, you might want at least seven days, but it's not required — just know they're gone for good once the window closes, though you'd still have your own copy on your computer.
Katey: Yeah, and I guess this isn't a trash-bin situation — it's not recoverable.
Brenton: Correct — for privacy and security reasons, it's not recoverable. Moving on — once the renaming's done, you can also download everything into one file, in alphabetical order, ready to save or send wherever needed. You can download or edit individual documents too, reclassify a document if needed, and use the three-dot menu for a deeper review, to report an issue, or to delete. We'd encourage everyone to use the report button if something looks off, even slightly — reports go to real people, not AI, and we can explain or fix things quickly.
While the payslips load — I've given one applicant two jobs here, a 9-to-5 and a casual weekend role — you can click through and flick between the different payslips within Doc Handler.
Katey: So essentially it's taken a bunch of oddly named documents, sorted and organised them, and now it's taking that a level further by sorting and organising the payslip income itself?
Brenton: Correct — it's effectively doing an OCR read, then categorising and labelling the income.
It works out and cross-references net-to-date, net and gross figures, super, names and ABNs across documents, and calculates income — but it goes further than a number, giving Jiffi-style analyst commentary you can discuss back and forth with, like an actual team member. For example, it flagged that all four payslips here were dated from the same month — not a real scenario, but it shows it's not just reading income, it's flagging things like 'you might have the wrong payslips' before you waste time building the application. It's also picked up two jobs instead of one, and flagged both ABNs as failing verification — again, hypothetical here, but in real life you'd confirm with the client, perhaps a recent re-registration.
You can click through to see which payslip it's referencing, and it's noticed two differing pay rates — possibly a pay rise, which you can confirm back in plain language, like 'confirmed pay rise from November at the higher rate', and it'll recalculate using the confirmed figure rather than defaulting to the lower one, flagging the gap between the two possible annual totals. Same process for the second job and the second applicant, which I'll skip through for time.
Then there's Compare & Push, based on every document read, not just income — going back to the earlier point, if you've got the documents and don't need to workshop the deal, there's no reason not to start with Doc Handler directly: upload, it renames everything, and — where before Pro you'd manually key it all in — now you don't have to. This example includes credit reports, an external fact find, ID, payslips and credit card statements, and it's even picked up an existing home loan from the credit report. Since it's a new scenario I'll leave everything ticked, though you can untick anything you don't want carried across, or push into an existing scenario instead.
It's flagging one applicant's income as unfinalised and a living-expenses warning, since we're not analysing those yet — that'll come with open banking once Fact Find and Doc Collection is live. After filling in the remaining income, everything here is prefilled from the actual documents — at this point I could export straight into a lender's calculator with no further manual work. Without this podcast slowing me down, that would have taken two to five minutes, and I'd have a shortlist of lenders I know are servicing, ready for the next stage.
Katey: Nice — and I guess that's really the goal: obviously this is a demo so we're talking through the detail, but once you're actually on a call, and say it's a refinance where you already have documents, you can run Doc Handler with what you've already got to prefill and workshop the scenario while you're still on the call.
Brenton: Yes, exactly — the less you have to ask the client for, the better. Our Doc Collector, part of the Fact Find, will also prefill a lot of the client-facing fact find itself — either using documents you already have, meaning less work for the client, or from documents the client uploads themselves, meaning they might barely have to fill anything in. Automating everything end to end is the ultimate goal.
So the final piece — Loan Submission Notes, also AI-generated and customisable with your own template. Get into the habit of adding summary notes — I'll do a brief one: existing client looking to purchase an investment property, using $200K equity from their owner-occupied home, good conduct on existing debts.
Katey: In a real scenario I'd add more detail than that, but for time — good conduct, there we go. Just checking, is there a voice-to-text option for notes too?
Brenton: Yes — you can either type, like I just did, or, which I'd recommend, add notes against almost anything in Quickli. For example, noting that a client's year-to-date income looks lower because she's only been in the role three months helps explain a figure the assessor would otherwise query, and gives the AI more to work with for richer, more personalised loan notes. You can voice-to-text into individual note fields, or add a summary note at the end — it's a personal preference.
Get into the habit of adding as many notes as possible. I'll add a random property address to show what flows through into the loan notes — Southport, Queensland — link that up, and add some income too. Based on everything from Doc Handler, this is all prefilled. When you're ready, you can generate the loan submission notes from a few different places — no right or wrong — and you'll get a summary first, which is your chance to add extra notes or mark anything as internal.
Internal notes just mean the AI won't surface them in the assessor-facing copy — for example, 'get Sarah's next payslip, she's done overtime' doesn't need to go to the assessor, so you'd mark that internal. You can upload more documents at this stage too. Then, generating the notes for a chosen lender, say Bank West — depending on how many notes and documents you've provided via Doc Handler, you might get six to fifteen prompts, more if you opt into the exit-strategy questions. The more information you provide upfront, the fewer prompts you'll need to answer manually, because the AI may have already answered them, or it might surface something you hadn't considered. If you opt into exit strategy, it pre-populates retirement-related questions — applicant ages, retirement timing, super balances — which I'll fill in quickly to show the generated notes.
Katey: And I guess this is really what's called context awareness — whether you're talking to an AI or working through a scenario, the more information available, the more you can validate the reasoning behind taking a client down a particular pathway.
Brenton: That's correct — which is why this first note was so brief; in a real scenario I'd add more detail, or use voice-to-text, because more context produces more extensive loan notes, and everyone — your assessor, your compliance notes, wherever you paste it — appreciates that depth, especially in this industry.
If you think the AI can't pick up on something, it can — for instance, since I selected a package product, it's asked whether the client wants an offset account, which I'll confirm. Whatever's entered gets commented on — so the more notes, the better. Because I entered a property address, and this is a purchase or pre-approval, you'll get commentary against requirements and objectives, additional comments (based on the brief note I added, though this section would be much larger with more detail), loan structure, product, applicant details, employment income, liabilities and living expenses, and CCR comments.
That last part's a bit more manual right now — I had one applicant's credit report, so it's commented on it, but not the other's, so you'll see a placeholder template there instead. You'll still need to source credit reports the way you currently do, but once Fact Find and Doc Collection is live, you'll be able to capture privacy consent and pull credit reports directly and quickly — one less manual step. Then there's exit strategy, risk commentary and recommendations. We can't dictate how brokers use these notes — some copy them straight to the assessor or compliance file, some go into settings and paste in their own required template alongside ours. I spoke to a broker recently whose brokerage template covered most but not all of their required points — the fix is simply to combine our template with theirs, so it covers everything in one pass rather than duplicating entries, since it all lives within Quickli anyway.
Katey: And I guess, as much as you're saying copy and paste, this is also reducing copy and paste — what's it called, the statement of... sorry, I'm not a broker!
Brenton: Statement of requirements — no worries! What I mean is: going from Jiffi through Doc Handler to loan notes, it's all captured, so you're not rewriting anything, especially with your own template — you can download it as a PDF or Word document, or ask for changes, which show up as tracked additions and deletions you can accept or reject, fully customisable like everything else in Quickli.
I've probably talked longer than I needed to, so I'll wrap up there, Katey, unless there's anything else to cover for your audience.
Katey: Not at this stage — and this is a good opportunity to say, if people have questions, they can pop them in the comments or reach out to you at Quickli directly. If someone wants to learn more or dig deeper into these features, where should they go?
Brenton: You can reach me directly at brenton@quickli.com.au. If you're already a Quickli user and need a fast response, our customer service team — bottom right of the screen — is the best bet; it's all real people, though they're fast enough that people often assume it's AI, and you'll typically get a response within a few minutes. If you're brand new, just search inquickli.com.au and you can start a free trial — everyone gets 21 days free with no credit card required. I'd be shocked if you didn't like it, but if you don't, there's nothing to cancel — after 21 days you simply won't be charged. If you've signed up before and it says you're not eligible for another trial, but it's been a while, reach out to us — in most cases we can override that, since so much has changed, and continues to change, over the last five years.
Katey: Perfect — thank you very much for jumping on the pod and showing us how to use Quickli.
Brenton: No worries — thank you for having me.
