Modern accounting firms are drowning in conversations. Every client meeting, board session and planning chat creates 20–40 minutes of admin that someone has to write up, file and action. That’s the gap Trent McLaren is trying to close with Vinyl, an AI note‑taking and meeting intelligence tool built specifically for accountants and bookkeepers.
In this episode of the Pinch podcast, Trent talks through why client conversations are the most under-used data in a firm, how Vinyl went from idea to hundreds of paying firms in under two years, and what he’s learned about using AI inside both an agency and a startup. For business owners and advisors, it’s a practical look at turning real‑world conversations into better decisions, less admin and stronger client relationships.
AI note-taking for accountants is about more than saving time on meeting minutes. AI note-taking lets firms capture, structure and reuse client conversations so they can deliver better advice, spot new opportunities and reduce unbilled admin. Instead of partners spending 30–40 minutes on a file note after every meeting, the AI records, transcribes and summarises the call in minutes.
Trent’s starting point is simple: every meeting already happens, and every meeting already needs a file note. That makes it the perfect AI use case. Vinyl joins the call, produces a summary from an accountant’s perspective (not a generic “here’s what you talked about” recap), and pushes that into practice management systems so follow‑ups and documentation are handled before anyone has left the room.
Because Vinyl is built for accounting, it understands context that general tools miss. It can tell the difference between a Chrome extension and a tax extension, and it recognises abbreviations like FBT, GST or ATO lodgements without getting confused. That may sound small, but it’s the difference between a rough transcript and something a reviewer would actually accept as a file note.
The real power comes from what Trent calls the intelligence layer. Vinyl doesn’t just summarise; it analyses what was said and what wasn’t. If a client sounds unsure about their budget or cash flow, the system can flag low financial literacy at the client level. If a question about fringe benefits tax is left hanging, it can tag an FBT opportunity for the advisor to follow up. Over time, firms can review 40 hours of meetings in a single query, pull out trends, and see where they’re leaving revenue or advice on the table.
Vinyl started as a straightforward idea: take the meeting assistant concept that already existed in tools like Fireflies and make it work properly for accounting and bookkeeping firms. The team had been using a general tool in their own agency and kept running into the same issue — the summaries were fine, but not built for compliance, tax or advisory work.
Instead of trying to serve everyone, Trent and his co-founders went all‑in on a niche. Vinyl records meetings, transcribes them, creates accountant‑ready summaries and then pushes the outcomes into the systems firms already live in. That might be practice management, proposals, scope documents or internal file notes.
From there, they added what many firms quietly wish existed: a way to go from meeting to actual work without retyping everything. Post‑meeting, users can ask Vinyl to draft a next‑steps email, a scope of work or even board minutes. What used to take two hours after a board meeting now takes seconds for some customers.
As AI capabilities improved, they layered on a chat-style assistant that already knows the context of every recorded meeting. An accountant can ask, “Look at this client for the last six months and tell me the main trends,” and get a coherent answer drawn from dozens of conversations. For firms dealing with cash flow questions, pricing discomfort or recurring compliance issues, that sort of conversational data is gold.
Importantly, Vinyl is designed to act, not just report. Trent is clear that firms don’t need yet another dashboard showing ten new “signals”. They need the right follow‑up to happen. That might be a drafted proposal in tools like Socket based on what was actually discussed in the meeting, a reminder to talk through cash flow scenarios, or a prompt to check whether an ATO deadline has been met.
On paper, Vinyl looks like an overnight success: more than 1,500 beta users before launch, 100 paying customers on day one of billing, and hundreds of firms now joining from Australia, the UK, North America and South Africa. Underneath that is more than a decade of very specific experience.
Trent describes his career in accounting tech as “the best apprenticeship ever”. He’s been an SDR making 50–60 calls a day, a field rep selling QuickBooks around regional New South Wales, and a public speaker on the global conference circuit. Before Vinyl, he built Journey, a specialist agency helping accounting software vendors launch and grow in markets like Australia, the UK and the US.
That background meant Vinyl wasn’t a cold start. The team knew the problems firms complained about, the events that mattered, the partners they needed, and the channels where accountants actually pay attention. They launched Vinyl in beta globally, ran invite‑only access to build demand, and showed up at industry events instead of hiding behind a stand.
They also invested their own cash. The first version of Vinyl cost somewhere in the range of $60–70k to build, long before external funding arrived. Early on, most of the code was written the old‑fashioned way. As AI coding tools like Cursor and Lovable improved, they progressively shifted until around 90% of new code is now AI‑written, guided by human developers.
For new founders, Trent’s message is blunt: product is no longer the moat it used to be. With modern AI tools, many people can build an interface that “works”. The real differentiator is still distribution — clear strategy, a defined customer, consistent channels and the ability to sell. It’s why Journey still exists alongside Vinyl: one business pays the rent and keeps the team in market, while the other pushes the product frontier.
If there’s a pattern across Trent’s work, it’s this: use AI to remove drudgery, not relationships. The most effective AI use cases in firms are the ones that reduce manual admin so humans have more time for high‑value client conversations and better business decisions. That applies whether you’re running an accounting practice, a tech startup or working inside a large corporate.
Inside Journey and Vinyl, the team records almost everything by default, then lets AI tools ingest and index the transcripts. They use that data to brief creative teams on video clips, build interactive research reports, and even create internal sales tools that combine product usage and billing data. Instead of bolting on generic automation, they look for places where conversations and workflows already exist and ask, “What would this look like if AI did the heavy lifting?”
For corporate teams, Trent’s advice starts with governance. Staff shouldn’t be pasting sensitive customer data into free AI tools, but they also shouldn’t be blocked from experimenting. Leaders need to provide safe, sanctioned AI environments and a clear policy, then actively encourage the team to try new projects — whether that’s a better internal dashboard, a smarter reporting pack or a more useful way to share research.
He also flags a new form of competition: your customers now have access to the same AI building blocks you do. In some cases, they’ll try to replace off‑the‑shelf tools with something they’ve built themselves. That makes it even more important to deliver genuine value, not just features. Products that save real time, improve cash flow visibility or make compliance less painful are harder to replace than ones that simply move data from A to B.
For firms thinking about cash flow and payment processes, conversational data is an important piece of the puzzle. Every time a client expresses uncertainty about their budget, payment terms or invoice timing, that’s a signal. AI meeting assistants can flag those weak spots so advisors — and tools like Pinch — can help businesses match payment options, billing cycles and automation to how their customers actually behave, not just what’s on the spreadsheet.
When AI is used this way, it doesn’t replace the relationship. It makes the human moments count more.
The through-line in Trent McLaren’s story is that conversations are assets. Accounting firms have spent years obsessing over financial data while letting the richest context — what clients say in meetings — disappear into memory, scattered notes and inboxes.
AI note-taking tools like Vinyl change that equation. They turn every meeting into structured data, surface missed opportunities and trim hours of unbilled admin from the week. Combined with thoughtful use of AI in coding, reporting and internal tools, they give firms the breathing room to focus on relationships, not retyping.
For business owners and advisors, the takeaway is straightforward. Record the important conversations. Use AI to capture and connect them. Then act on what your clients are really telling you — about their cash flow, their systems and the support they need.
That’s where the next wave of better business processes — and more confident cash flow — is likely to come from.
Transcript lightly edited for clarity, readability and blog formatting. Timestamps, filler words and auto-caption artefacts have been removed where possible while preserving the meaning and flow of the original conversation.
Trent: Yeah. So we can load 40 hours’ worth of meetings into a single prompt and then say, “Pick out all the trends. Tell me what my sales playbook is. Tell me how I could have made more money.” And it will analyse every single word and say, “Hey, this is what we’ve heard in the 40 hours’ worth of meetings you just gave us.”
Joe: Welcome to PMID Episode Five. Today, I’m joined by a very special guest, Trent McLaren — the crown prince of the accounting tech industry. But before we have a chat with Trent, we’ll go through the formalities. We’re on YouTube, Amazon, Spotify and Apple Podcasts. Like, follow and subscribe to us on all of those channels. We’ll be recording heaps more episodes and heaps more content. Trent, thanks for coming on.
Trent: Joe, how are you?
Joe: Great. Good to see you.
Trent: Thanks for having me. What a fantastic setup we’ve got here today. I’m enjoying it.
Joe: Yeah, we’re here at the ICC. We’ve just found a little pocket in the corner because we’re here for ABE. What brings you to ABE?
Trent: Accounting Business Expo in my hometown of Sydney — it’s hard to pass it up. A lot of people have been asking me what brings me here because we don’t have a stand or a stall. Actually, it’s for the people. It’s for this. I just love catching up with people. We spend so much time online engaging, commenting and posting, but then you meet someone and they’re like, “Oh, are you Trent?” Or you meet an Ian or a Shay or whoever, and you actually get to connect in real life. It’s fun. I like it. It lets you build that extra relationship a little bit faster because of those digital relationships that we already have.
Joe: Yeah, absolutely. We’re here with Annature to launch our partnership, so I think we’ll talk a lot about partnerships and that kind of stuff. You’ve obviously helped us a lot with our strategy around that. But before we get into too much nitty-gritty, tell us about you. A lot of people know you in the accounting industry, but there’ll be people listening to this who are tech founders, people in the payments industry and general small businesses. Start from the beginning. What’s the Trent McLaren origin story?
Trent: I’ll tell you the truncated version — the succinct version. This is part one of a three-part series. No, I’m kidding. I’ve been working in accounting tech for nearly 14 years. I always say I’ve had the best apprenticeship ever when it comes to accounting tech. I’ve done every job you need to do within a software organisation on the commercial side. SDR, making 50 or 60 phone calls every day, being in my car selling QuickBooks licences around New South Wales and the country. I’ve worked in other companies, travelled the world, attended different conferences, done public speaking — the works. I’ve probably got more T-shirts and hats than most people when it comes to software companies.
My main focus right now is a company called Vinyl. We’re an AI note-taking assistant for accounting and bookkeeping professionals. We record your meeting, transcribe it, summarise it and push that data into your practice management system. We’re also about to go really heavily into the intelligence layer that sits on top of that, which is super fun. It’s a big elevation of where we’re headed, but also where I think AI is going to take a lot of products this year.
Joe: Let’s talk about it. What’s the intelligence layer? What do you mean by that?
Trent: When you record a conversation at the moment, we’ve got a transcript and we use that transcript to create a meeting summary. But we’re doing that from the point of view of an accountant or bookkeeper. We’re calling out all the key things that are important to you. We’ve picked up all the abbreviations. We understand FBT, VAT, GST, lodgement with the ATO. These are things that not all AI systems pick up because they’re horizontal by nature.
If I talk to you about an extension I need in my business, am I talking about a Chrome extension or a tax extension? Knowing the context of what I’d be talking about if I’m an accountant isn’t necessarily programmed into a lot of those horizontal layers.
We’re in the process of developing what we call “signals”. The idea is that we can review a transcript, index and tag that conversation and identify phrasing that creates opportunities you might have missed. Things that you said or didn’t say as the client, or that I said or didn’t say, which we can now pick up and flag.
We might say, “Hey, this sounds like an FBT opportunity you need to talk to your client about.” Or, “This sounds like revenue you could be picking up from your client that you’ve missed or haven’t answered properly.”
If I ask my client how confident they are with their budget or cash flow and they give me a mixed or very small answer, our AI can detect that the answer to this question was quite low. They’re not feeling particularly confident or engaged with that service line. That might be something you want to talk about. But it’s also something we can store on file against the client record. What’s their financial literacy like? Well, based on what we’ve indexed and observed in the transcript, we’re saying it’s low right now.
That’s the intelligence layer we’re talking about. It’s analysing the words that were said and not said to help the firm ask, “What are we missing?” When you’re in a conversation, you’ve got tunnel vision. I’m focused on you. You’re focused on me. We’re not thinking about all the other things. This intelligence layer is trying to pick up those other things.
Joe: The subtle cues. Are you pushing that data into other locations? Does Vinyl integrate with CRM software and things like that? Or are you also building functionality that helps hand that information into workflows?
Trent: The important part is that what we need less of is another dashboard saying, “Hey, we picked up 10 signals in this call.” It’s the action that needs to happen next. That’s where we’re getting super excited.
Right now, in our system, you can record the meeting, see the summary and start asking it to draft the next-steps email. If I wanted, I could ask it to draft a scope of work or advice or whatever it may be. Our AI chat-writing function will actually do that within the system.
Joe: As in, if you’re having a conversation, can you stop and talk to the AI while it’s recording you?
Trent: It’s all post-meeting. We don’t do any live-aid type stuff yet. The idea that you could ask a question live and have it come back to you is something we’re exploring. I mean more like having a GPT-type function within our app. It’s a chat-based function that already has all the context.
If you want to say, “Look at this client for the last six months. Tell me all the trends that have happened in those meetings,” we can do that now. We can load 40 hours’ worth of meetings into a single prompt and say, “Pick out all the trends. Tell me what my sales playbook is. Tell me how I could have made more money.” And it will analyse every single word and say, “This is what we’ve heard in the 40 hours’ worth of meetings you just gave us.”
Joe: So Vinyl has an AI assistant component that you can talk to which basically aggregates all the information it’s collecting?
Trent: One hundred per cent. We’ve had a meeting. Post-meeting, I want to draft my next steps and my actions. I might want to create a file note as an accountant or a bookkeeper. We do all of that in seconds. We have people who record board meetings that would normally take them two hours after the meeting to produce the board minutes. We’re now producing those board minutes in seconds. That’s wild for those people.