Let’s chat about that financial complaint you gave up on

Let’s start with introductions—who are you, and what are you doing?

Dan: I’m Dan Naratnam, co-founder and—to give it an elaborate title—CEO of Womby.

Jon: And I’m Jon Manning, CTO. Though it depends on the day and which hats we’re wearing. Really, we’re just two co-founders.

Tell me about Womby.

Dan: Womby is named after the wombat. We were looking for a spirit animal and a brand that would suit the kinds of things we're doing. A strong rear end, fast when needed and great at digging. 

Today, publicly, it’s a platform where you can tell us about a complaint you have with your financial services provider—a bank, insurer, superannuation fund or investment fund. We only ask you about the complaint itself, nothing about you personally; that’s not important to us. We use AI to match those details against past complaints and past decisions from the financial ombudsman, then we retrieve those decisions, summarise them for consumers, and link out to the original documents.

It started from the feeling of not being alone in a complaint. Knowing that someone else has had a similar issue is a good first step—your situation probably isn’t unique. It might be in some cases, but broadly, people complain about similar things. Complaints have a similar shape. We just wanted to show people those similar shapes, and use AI to make that simpler.

What does it actually look like for a member of the public to engage with Womby? What situation would I be in when I come to you?

Dan: Go back to the origin. I had a complaint with my bank—I’m a cradle-to-grave customer of one of Australia’s largest banks, and this was the first time I’d ever had an issue. I assumed it would be straightforward, but it was long-winded, arduous and hard to navigate. I wasn’t even sure how to frame it the right way.

The front end of banks is highly digitised through their apps, but the complaints side is the poor cousin—you notice that pretty quickly, it’s also inconsistent across firms. 

So today, if you want to engage with Womby, it’s a website. You can use it on your phone or any device—it’s responsive and device-agnostic—and you interact by filling out a form. We ask some basic questions, including a couple of things that matter to us: whether you’ve already complained, and how you’re feeling. We put some emojis in there for the mood check, and most people are unhappy. No one has ever hit the positive service-experience emoji. Very few say they’re optimistic about getting it resolved. It's mostly “I’m enraged” or “I’m frustrated with the process.” That’s valuable for us to know.

It started from the feeling of not being alone in a complaint.

There’s clearly a personal experience behind this. How did the two of you come together—it’s one thing to have a frustrating moment, another to start a business over it?

Dan: Jon and I go way back. We’d worked together and had been talking for a while about wanting to work in the same place again. A couple of years ago—we were in separate companies—we were right on the edge of AI becoming mainstream, and we were both looking for real use cases for it.

We’d talk about what we were each doing privately, and how mind-blowing it was that this tool was suddenly so accessible. A lot of people around us were being cynical and disparaging about it. Initially that was me. When Jon started showing me what he was pushing it towards, I felt cautious for the first time ever. The one skill I really held — writing—suddenly felt commoditised. I was resentful that the thing I was good at was now available to everyone, because that’s what used to separate you. Of course that’s not really true—people are craving authentic words more than ever—but that was my reflection point.

My own issue was what got me in. I went through the complaint process and discovered there was an ombudsman for financial complaints. The problem was, I couldn’t find anything useful around outcomes—you’d end up on ProductReview, Reddit, forums or Facebook groups, and people there are polarised: happy if they got a good outcome, furious if they didn’t.

The ombudsman’s own site is extensive, with lots of resources, but it was tricky to navigate as a newcomer. The thing that stood out was the past decisions—the first time I’d ever seen a fully fledged account of a complaint: here’s the issue the person had, here’s how we made the decision, here are the rules that were followed. Ten pages, really well written and articulated. I found that exciting. But nobody we spoke to knew it existed.

So I called Jon. I said, I think there's something here—I think AI could really help get someone closer to the mark, rather than it being one more thing to check.

Jon: That’s roughly what I was saying too—AI could help with finding, cataloguing, filtering and summarising. It seemed fairly neutral because it’s public data. For our purposes it was a great experiment: let’s see how much data there is and how far back it goes. Let’s see if AI can really help in searching for things and matching the right files.

Dan: We did some very early proof-of-concept work—the classic “everyone accuses everyone of building a ChatGPT wrapper.” Well, of course. We wrapped a handful of PDFs, handpicked some determinations, dropped them in raw with no tuning, and just watched what it spat back.

These determinations — are they the case-study outputs of the complaints process?

Dan: They’re the end of the secondary dispute process. When you have a dispute with your financial firm, the first stage is IDR (internal dispute resolution). That’s the initial complaint you make, and firms have to follow a defined process to help you navigate it and reach an outcome. The process ends with a letter telling you the outcome, and in that letter they must tell you where your recourse is.

Your recourse is a secondary review with the Australian Financial Complaints Authority (AFCA). That’s the EDR stage: external dispute resolution. AFCA reviews whether the first process was handled properly. If the firm did it well, has the evidence, and the decision was right, AFCA can close it early at what they call registration and referral. If this isn’t the case, it moves to case management and the further this goes, the more it costs the firm - the service is free for consumers.  If an AFCA case manager fails to reach an outcome with both parties,  it goes to an ombudsman decision, the outcome of which we summarise for consumers. 

When you say “firm,” who are you referring to there?

Dan: A bank, a superannuation fund, a general insurer, a life insurer—that can include pet, car and home insurance—financial advice, ETFs, that sort of thing. Anyone managing your money who has an AFSL (Financial services license) So it’s broad, and there are so many problems in those spaces.

People ask what you can typically complain about—anything where you feel you’ve been impacted. And there’s a formal definition: firms follow ASIC’s rules on what actually counts as a complaint.

Dan: The core of what we do comes from understanding that many people don’t know how to complain, or don’t do it well enough. We pitch the idea of ‘How do you complain, better?’  These things often slip through as “product feedback.” The problem is the heightened emotion around being financially impacted—the way people express that often doesn’t come out as a complaint at all.

There’s a formula: this action took place, I was impacted in these ways—emotionally, financially, socially—and here’s what I'd like in return, whether that’s an apology, an explanation, money back, or a reversal. People tend not to do that, because they don’t know or the structural ask from a form is a blank text box - this is a common UX pattern. 

Jon: There’s also the “keep clicking next” behaviour. Once someone gets a response from their bank or insurer, all they’re scanning for is: is my thing solved or not? When it says “not in your favour,” I'd bet most people stop reading right there and never get to the part that says here are your next potential steps.

Dan: Because pursuing it takes time, effort and energy. You see it in the numbers, the drop-off from first complaint is huge. Roughly five million people complained last year, and a lot are missing from view. ASIC did a study back in 2019 that called these people “considerers”. They know they have an issue, but everything feels structurally stacked against them, so they don’t bother. That’s a lot of hidden financial harm. People have been hard done by and then set up to fail.

That power imbalance feels like a big part of this.

Dan: It is. There are free services out there—we have good, collegial relationships with the not-for-profit sector, and some are interested in what we're doing—but they’re massively overburdened. And financial hardship is now a reality for mainstream families. There is no longer a fringe of people doing it tough.

Five million complaints- is that the real number?

Dan: That’s the registered figure—people who actually called, emailed or otherwise made a complaint. ASIC has those statistics because firms have to report them. There’s likely some underreporting in there too; some firms registered no complaints in a year. That seems odd.  ASIC gives firms templates to respond to, but you assume some margin of error on what’s missing. And that’s before you count the people who just won’t get off the couch to do it.

So Jon and I asked: can we make this easier, give people a different perspective, reduce the friction? It’s never been about giving people false hope about the outcome. The purpose is to show that others have been impacted by these firms—and the power of AI is that within twelve pages of detail it can surface, say, “cases similar to yours succeeded when they had photographic evidence or an expert report, and worked against the consumer when they didn’t,” or when the insured hadn’t read the policy terms and the insurer was following them. AFCA isn’t a court—they administer the rules and arbitrate—so there are things people should be aware of, and AI has potential to expose that.

So the motivation seems to sit at the intersection of a genuine excitement about AI as a one-to-many enabler, your own frustrating experience, and a large body of public data—coming together to make a fairer space.

Dan: Our mission is to help create faster and fairer outcomes for consumers. There are two things there. One is speed: the IDR-to-EDR process can drag out—a super complaint might last half a year. It’s exhausting, AFCA is overburdened, and it takes brave, determined, resilient people to stick it out through the administrative burden and the emotional and financial toll. People don’t have time. Firms do—they have the resources, and it’s literally their job.

Our view is that if we can prepare the consumer at that first stage—quickly and easily, so they understand their rights—and go beyond the public data into the codes and standards so they understand the firm's obligations, then we can get people away from leaning on no-win-no-fee third parties who are effectively doing the same thing our AI might be able to do. Those are for-profit firms; they have to monetise the model, so they take a fee. We would like to think that model could come to an end. AI represents a succinct end to it—in the future, consumers can access, free or for a very small fee, the same information in the context of their dispute, the same thing a highly paid person would have interpreted for them. We want to make clear that humans are still needed here. That’s why there are many NFP firms helping provide agency to consumers lost in this process. We think that AI can play a role to help get people started the right way so that experts are left to deal with the gnarly issues. 

Jon: And to be clear, we haven’t yet monetised this. We run it privately, at a fair bit of personal cost. 

Dan: Minor to great discomfort, honestly. Everyone romanticises startups, but there’s a lot of personal sacrifice and investment for both of us, time and money. We’ve forgone income, and we’ve deliberately not sought third-party investment, because we don’t want the mission skewed by having to pay investors back.

We need to be very clear about that, because others in the industry might accuse us of doing something wrong. We offer a free service that collects no personal information. If someone wants to challenge that, they’re welcome to a conversation. We’ve collected no money and no personal data that can be monetised. The only thing we’ve collected is many financial complaints against firms.

Jon: That’s the important part: we’ve collected thousands of complaints across all the firms, so we can see the mistakes they’re making and the harm they’re causing. That’s hard to read. Disappointingly, but unsurprisingly, a lot of interest in us has come from the no-win no-fee firms. 

Dan: The ambulance chasers—and they came at us aggressively, because they could immediately see that what we have could help them scale their business. That’s validation. We don’t want their money, but thank you for confirming we have something that’s an existential threat to your business model—it’s our personal view that that business model needs more regulatory oversight.. We sometimes get blowback for what we do, but nobody’s giving feedback on those firms.

From our conversations over the last 2 years, there’s acknowledgement that no-win no-fee complaint businesses reflect financial product and complaint process failures.  At least there’s acknowledgement of the gaps in the system that lets them exist. They’re parasites.

We offer a free service that collects no personal information. If someone wants to challenge that, they’re welcome to a conversation. We’ve collected no money and no personal data that can be monetised. The only thing we’ve collected is many financial complaints against firms.

So would it be fair to say you're consumer-aligned, and always intend to be?

Dan: We can actually operate on both sides. Asymmetrical data access is a problem for firms as much as it is for consumers—but let me go back to how we started. We spent six to eight months on what I'd call “professional fucking around”: UX research, downloading and reading a lot of papers, a lot of talking. Then, over December, it was Andrew Walker—an ex-colleague and one-time successful founder—who kept saying, “Why don’t you guys just build the thing? You’re talking a lot.” I didn’t listen at first.

Jon: We’d proven the concept could work, but we were still spinning our wheels, thinking we’d have to find an expert to build it. We’d tried the ChatGPT wrapper, and I said we were going to have to do proper AI—actually build it. And I didn’t even really know what that was. 

Dan: Jon and I are long-term data and technology people, but you don’t get exposed to building this kind of thing inside a big organisation—you have to be privileged to be part of that. So Jon just went crazy over the break and built something, then sent me a screenshot.

I asked, “What’s this, a mockup?” He said, “No — it’s the thing.” I said, “What thing?” It had typed in a thing, searched the thing, and come back with a thing. And I was like—that’s the thing. The thing’s doing a thing. That was December 2024. We were live with a website and a working product by March 2025.

It’s a different world. Our shield has always been that we can go public because there’s no potential for harm to consumers—it’s free, it’s anonymised. What’s the harm? Well, the industry can find a lot of harm in it, because it’s highly disruptive. ASIC and others report at an abstracted, statistical level—stats are great but they tell you nothing in detail, and that obfuscates the useful detail. We’re a collection point.

Jon: One thing we find really interesting about Womby is that it’s a different take on accessibility—whether that’s digital literacy, or just the comfort of sitting down and finding that information online. We all have people in our lives who’d struggle with that. Dan’s a senior technology person with a professional background and still found it hard to find. So there’s something here about levelling the playing field.

Dan: It’s also the financial literacy of synthesising a lot of information quickly, and AI unlocks that. Using Womby publicly takes about four minutes, and you come away with fifty past decisions synthesised for you and one-click access to each relevant one. What you do with it from there is up to you. We are extending this to codes of practice, laws,  rules and standards.  Our strategic advisor Ron Arnold will often refer to this as “the AI powered consumer”. 

Being free, with no registration and public use, has given us a lot of data on how people complain and what they type. The tricky part is that people tell us too much—there’s an assumption that more detail makes it better, but that’s not true. If the context window is too big, you can’t get an effective semantic match.

How do you counter that—is it a UX thing?

Jon: It’s totally a UX thing. That was amateur hour for us early on—we didn’t know it. We’ve done three revisions of the product, and we’re on another one now. It’s all in the back-end mechanics: taking what someone says and refining it down to the core parts we need. We don’t need their life story. We’ve spent a lot of time learning how people complain to test our designs, listening to the participants including regulators and building trust. 

Dan: And we've heard that firsthand. Heads of disputes at some of the biggest financial firms in Australia have told us people literally tell them their life stories or get ChatGPT to write the tome for them. It happens—you feel wronged, you’re upset, and the natural instinct is “let me tell you how bad this is.” It’s almost cathartic to unload. But people are committing too much information, or information we don’t want to hold, so we had to design around it.

The easy questions are the ones we can bucket—who was it, roughly when did it happen, how much were you owed, what was the specific problem. The hard part is the narrative. That’s the real accessibility challenge: telling and shaping the story. There’s a way we want people to express it—ideally the way Jon wants it for the AI—and you’re constrained by a text box. We even added a microphone so people can dictate, but we don’t hold those recordings. We’ve been learning all along about the best experience for consumers to elicit their complaint and the best data structure to capture it. 

On the topic of copyright: we've taken the decisions from previous complaint processes and created our own AI summaries and metadata to find them. We build everything on our own data. 

How many of those are there?

Dan: More than 32,000 and counting, going all the way back. From an interpretation standpoint—rules, laws, codes, standards—the people making those decisions change over the years, so there’s variability, and there’s always an argument about consistency. Everyone raises the same two things with us: hallucination and consistency.

We don’t think consistency is our problem. This is public data. Taking into account that rules, codes, products and policies change over time, whether one decision is consistent with another is for the firm and the consumer to work out—but that’s not a reason to never show it. The explainability is always there but it’s common sense to look at the most recent decisions to see how rules are being applied today. While everyone’s circumstances are unique, we are interested in their shape. 

Jon: On the engineering side, that variability was an early issue. Dan asked why we were getting variable results, and I re-engineered it. We needed to minimise what the AI does—down to retrieval, plus summarising the person’s issue in the context of what’s retrieved. Initially we were asking AI to do too much of the summarising, and that over-analysis is where the variability crept in.

So winding it back is a big part of getting a consistent result. How do you resist the natural pull to go deeper?

Jon: By applying it at the right place and time. Rather than handing the AI the raw complaint and asking it to do step one, step two, step three, four—all these call-and-response chains—we just extract the key pieces of information related to a case or determination up front. Once we’ve got that and stored it, all we really need to know is: which of the 30,000-plus cases are similar to this person’s complaint? Then we retrieve what we need. We’ve designed with our own costs in mind and optimised everywhere. 

Who is the typical user?

Dan: That’s the tricky part, and we haven’t fully cracked it, because we ask for no demographic data. We’re starting to move towards wanting to know a bit more about who’s using Womby and how much value they get. We’re exploring a few user journeys. We are often thinking of the value exchange for a consumer in asking for personal information vs the outputs we provide. We always want to give more than we get. 

Part of the monetisation path is: it’s great that we’ve given people insight, but how do we help them draft that crisp, slop free complaint letter—get them a step closer to actually complaining? We’ve validated with a number of groups that firms themselves want clear, precise, concise, legitimate arguments. They’re going to read it and want to know what to do.

When you say “firms,” you mean the banks and insurers the complaints are made to—so it helps them too, given the overhead of managing wayward or emotional complaints?

Dan: Right. There’s always someone in your circle who’s a lawyer, and people go, “I feel like I’ve been screwed over here, help me.” That lawyer does exactly what we can do at scale: check the codes and obligations, ask about the circumstances, apply the rules, and draft with precision. We want to make that available. And to be clear, these aren’t legal letters—they’re complaints.

The lawyer analogy is a nice one—it’s about a privileged social connection: knowing someone who understands the system.

Dan: The extension of that—and this is where we’re heading—is an assumption we now make: firms are going to read that letter with an AI agent.

We assume eventually, it’ll be read  by AI first, so we structure it accordingly. If the firm codes its AI to read according to the rules it has to follow, then we write according to those same rules—and we make sure it never slips into the feedback bucket.

Tell me more about the feedback bucket.

Dan: If a firm’s AI is reading the letter, it’s looking for a specific shape: “I had a problem on this date, I was impacted in these ways, and I’d like this in return.” That’s the ASIC RG 271 definition—there’s a formal definition for a complaint. If your letter or the words on your phone call or your social media post doesn’t meet it, it probably won’t be treated as a complaint; it gets treated as feedback. The clarity you provide also helps the AI route it to the right team. You’re never prompted for these components. 

It’s a bit like when everyone had Foxtel. You’d call to complain about the price, they’d say there’s nothing they could do, and only when you said “I’m going to cancel—put me through to the cancellation team” would you get to retentions and a discount. How your issue is triaged completely dictates the outcome.

So if you understand how the for-profit machine works, it’s that they don’t want every letter going to the complaints team.

Dan: 100%. But they do need to know where the drop-off point is, and they’ve validated that AFCA decisions are the ones they watch for. Complaint identification persists as an issue for firms, particularly as the preferred channel choice is the phone. For your audience: there’s a cost to the firm every time something goes to AFCA, borne by the firm. The further the consumer goes, the more it costs. The public decisions in Womby are a reflection of a determined consumer and an equally stubborn provider. We wonder if firms will sometimes stay all the way to a public panel decision from the ombudsman to establish a public decision marker. Jon and I have read many decisions and wondered why a firm didn’t just make a discretionary and private decision to settle much sooner. Surely that would be a better outcome?

So it’s not a precedent. AFCA insists it’s not a precedent. Everything on their website says “these are not precedent—but they’re indicative of how AFCA may resolve a similar issue.” All we’ll say is the information is really useful and the most recent decisions are a public, in context application of rules which industry participants find useful. 

So we rely on that, because the consumers or their advocate’s ability to surface it is a form of leverage—and consumers don’t have leverage, or they have to pay for it.

Dan: In our updated consumer journey, we are designing a way to point people to free resources. There are financial counselling firms in every state, national debt helplines. The problem might be bigger than the complaint—“I’m going under, I have debt, I need to speak to someone.” That’s almost a concierge moment: we point or pass people along first. You may not need us. And if you’re certain you do want to complain, then sure—you can write a letter, and that might cost a very small fee.

That’s what we’d ask people to pay for: not access to data, but drafting. We think that’s worth paying for, because we have to run a service—we can’t pay for the AI and hosting and particularly the security posture forever while holding down private jobs. We’re looking at grants and other creative ways to fund it, but that’s taking time. 

How did the second product come about?

Dan: Firms indicated the data was valuable to them, though they’re on their own AI path. Earlier this year, in January, we launched a different product called Burrow—a non-AI search for firms to find these disputes, sort them and collaborate on them.

When we spoke to firms, they said they had senior people who just know the decision numbers—“Decision 657, or 358, is one to look at”—knowledge held by individuals who’ve been there long enough, or kept in an Excel spreadsheet. I thought, why not make that a proper tool? Let them search, bookmark, save searches, see who bookmarked what, and see the latest decisions—because there’s currently no mechanism to see what the newest decisions are. We’ve been capturing all of that anyway.

Firms were interested, but the corporate sector is on its own journey and working with early stage startups isn’t easy for legacy organisations. So we’re still finding out who our ultimate customer will be—and the most interest has actually come from the not-for-profit sector, because we came to them without an AI product.

That means financial counselling firms, legal support centres—anyone giving agency to a person around an individual complaint. It’s a good alignment for us, because it helps more people more often. The sophisticated ones look at it and immediately get it, because we designed it knowing that the job is hard. The sector is exploring AI on its own terms. We think that any AI offering from us to these firms will come with trust and utility from the basics. 

Some well-known counselling firms already have Word templates for writing a complaint letter. They’re trying, but we can give them better execution. Funding for this kind of software is hard for them to access, and we’re in the rare position of being able to build it independently for everyone. There’s a real possibility to overflow to them—give them drafting tools, case management, ways to pull things together.

The trust bridge is being transparent: “Here’s some data and software for something we know is difficult for the sector, we’re still working out the funding model, do you want to try it?” We’ve talked about a circular model—if industry pays a licence fee, we could offer it to the NFP sector for less.  We’re not at commercialisation yet; we’re running experiments and talking to participants.

With a magic wand, what would a great outcome look like over the next twelve to eighteen months?

Jon: Ideally, a funding model that lets us keep offering the consumer side for free. From a product perspective, we’d love to see a consistent way for consumers to complain across all financial products. Right now, that experience varies from firm to firm. You get a real sense of how important this is by how easy or hard it is to find the complaint form or email address. If you go backwards from the type of data that would identify systemic issues, you start at consistent data capture across the sector. 

Dan: Exactly, I’d like to see Womby play a role in a national system that collects disputes data at a level valuable to the whole system. AFCA decisions are one thing, but they’re at the end of the chain. It’s everything at the top—the products impacting people, and the outcomes largely hidden from view, held only by the person affected or the firm.

We think there’s a way to do that through the organisations helping consumers, in a way that protects the consumer’s identity and focuses only on: this was the problem, this was the outcome. Doing that at a national level would allow sharing between agencies and point to systemic problems—holding firms to account for products that aren’t working.

Right now there’s no detailed data sharing whatsoever around disputes, and market concentration is high, so firms know which of their products are problematic. But we’ve never seen a firm come forward and say, “We changed these policy terms because we knew they were creating harm for our consumers.” We think better transparent data and incremental changes are needed. It shouldn’t take a royal commission. Surely some simple, publicly available data surfacing trends could help—“this is the worst credit card in Australia, evidenced by the number of complaints from these states—fix it.” Follow that through, and you start to get better financial products. We see admirable reporting from AFCA on systemic issues, but ideally this capture needs to happen much further upstream. 

Jon: So less problematic products, and transparent data sharing between the organisations helping consumers.

Dan: So if a firm is supporting a consumer in Sydney seeking debt relief on a loan, and someone in Perth is dealing with the exact same thing, right now neither knows the other exists. With shared data they could say, “I've seen this—here are the circumstances, here’s what happened.” They don't need to know who the person was—their gender, age, ethnicity, none of it. They just need to know it’s verified by a real person on the other side of the country. Then they can hold the bank to account: “I’m aware of this case.” There is fog on decision discretion around disputes. Today firms are able to rely on consumers operating in the fog. If having better access to data means solving faster and fairer, we’re all for it.

We have great advisors—industry experts who help us on the regulatory and strategic side, based in Sydney. So we’re not on our own. They’re aligned with us on going slow, making a market. Like us,  they’re very pro-consumer, but in parallel they understand the very real challenges faced by firms to handle complaints effectively. Their wisdom is grounding for us. They have worked across reserve banks, national regulatory agencies and the largest insurance firms in the country. They are credible career-long experts and wayfinders across strategy, rules and operations. Being able to seek counsel from industry stalwarts and experts who are passionate about your business is a privilege for a couple of guys on the wrong side of the country.

Jon: Having people of that calibre align with our mission and how we’re trying to help people was a real boost to us late last year. 

Dan: It’s that first moment of delight when they see Womby and go, “This is amazing—how do I help?” Ron said to me, “I’ve been waiting for someone to do this. I’m really excited.” 

There's nothing nefarious here—our bank accounts prove it, and our product collects innocuous information. We’re taking the right steps, working with the right groups, entrenched in the idea of advancing consumers. And that doesn’t have to come at the cost of firms. If firms make fair decisions sooner, they retain the customer. The whole point is retention, and we play into that. We want to see much less needless escalation to AFCA because a fair outcome was reached sooner.

There is fog on decision discretion around disputes. Today firms are able to rely on consumers operating in the fog. If having better access to data means solving faster and fairer, we’re all for it.

There’s a genuine win-win here, isn’t there—less time administrating, better product outcomes.

Dan: Fewer complaints, better for everyone. You’ll end up with a better NPS if you solve a problem than if the customer just churns. Australians don’t stand around a BBQ talking about their insurer being awarded by CANSTAR 5 years in a row. They talk about how their Insurance company solved their complaint quickly and easily.  People need to stay open to new ideas about how this happens. In a world where every firm is making deep investment in AI, nobody’s doing it the right way for consumers. That’s the path we’re trying to take—do it the right way, follow industry standards, and earn the industry-backed acknowledgement that says, “These guys are doing the right thing, and this is the best way to complain.”

Thank you both.