[Emerging Tech Series] Transforming Financial Advisors’ Workflows With AI Agents With Chirag Gandhi

Chirag Gandhi is the Co-founder and CEO of Mili, an AI-powered platform that automates administrative workflows for financial advisors, helping them save time and focus on client relationships. Under his leadership, Mili has raised funding and expanded its enterprise AI platform in the wealth management space. Chirag previously worked in venture capital and growth equity, where he helped raise and manage significant investment funds, gaining deep experience in financial services. Before founding Mili, he held roles in investment firms and served as an angel investor supporting early-stage startups. 

Here’s a glimpse of what you’ll learn:

• [2:22] Chirag Gandhi discusses how Mili’s AI agents eliminate administrative work for financial advisors.

• [4:02] Building AI specifically for wealth management and financial services

• [5:50] Why customization matters for advisor workflows and tech stacks

• [9:10] Real-time AI note-taking without bots or recordings

• [10:30] Preventing hallucinations, bias, and data leakage in AI

• [13:39] Chirag shares the long-term vision for agentic AI beyond meeting notes

• [15:34] The three pillars of AI agents: Ask, Automate, Alert

• [19:42] What AI should not do in financial services

• [22:06] Build versus buy decisions when creating AI agents internally

• [27:27] How Mili measures and protects against AI bias and drift
  

In this episode…

Financial advisors are overwhelmed by administrative work, fragmented systems, and rising expectations to adopt AI — without compromising compliance or client trust. Many teams experiment with AI tools, only to find they don’t fit real workflows or create new risks. How can firms use AI to save time and scale impact without losing control?

Chirag Gandhi, a fintech entrepreneur and AI systems expert, explains how advisory firms can apply AI in practical, compliant ways by focusing on workflow-specific automation. He emphasizes starting with meetings as a data source, using real-time note structuring, and integrating deeply with existing systems rather than layering generic tools. Chirag also shares why safe design choices, human-in-the-loop oversight, and clarity on build-versus-buy decisions are critical to making AI actually work in regulated environments.

In this episode of The Customer Wins, Richard Walker interviews Chirag Gandhi, Co-founder and CEO of Mili, about applying agentic AI in financial services. Chirag discusses designing compliant AI systems, deciding when to build versus buy AI tools, and how agents can automate tasks while preserving human relationships.

Resources Mentioned in this episode

• Quik!

• Richard Walker on LinkedIn

• Chirag Gandhi on LinkedIn

• Mili

• "[Emerging Tech] Eliminating Data Complexity for Growth With Mark Ovaska" on The Customer Wins

• "[Emerging Tech] Protecting Businesses From Privacy Lawsuits With Richart Ruddie" on The Customer Wins

• "[Emerging Tech] Reimagining CRM With AI for Advisors With Thomas Clawson" on The Customer Wins

• "Empowering Advisors With the Gift of Time With Rich Whalen" on The Customer Wins

• Rich Whalen on LinkedIn

Quotable Moments:

• “We are building AI agents that can streamline operational workflows for wealth advisors.”

• “You could have two advice firms separated by a few blocks in New York.”

• “We had a very cautious approach to deploying safe, practical, and compliant AI.”

• “The space between the advisor and the client is pristine.”

• “With agents, you should be able to ask, automate, and alert.”

Action Steps:

1. Start AI adoption with core workflows like meetings: Meetings are a consistent source of high-value data across advisory firms, and optimizing them first creates immediate time savings and better downstream automation.

2. Design AI with compliance built in from the start: Regulated industries cannot afford experimentation without guardrails, so embedding privacy, consent, and human oversight early prevents risk and builds trust.

3. Customize AI to existing systems instead of forcing new ones: Advisors use deeply personalized CRMs and workflows, and tailoring AI to fit those environments increases adoption and long-term value.

4. Keep humans in the loop for client-facing outcomes: AI can draft, organize, and recommend while final human review preserves relationships, accountability, and client trust.

5. Evaluate build vs. buy decisions honestly: Building AI internally requires sustained investment in data, talent, and maintenance, while partnering can accelerate impact and preserve focus on core strengths.

   
   

Sponsor for this episode...

This is brought to you by Quik!

At Quik!, we provide forms automation and management solutions for companies seeking to maximize their potential productivity.

Using our FormXtract API, you can submit your completed forms and get clean, context-rich data that is 99.9% accurate.

Our vision is to become the leading forms automation company by making paperwork the easiest part of every transaction.

Meanwhile, our mission is to help the top firms in the financial industry raise their bottom line by streamlining the customer experience with automated, convenient solutions.

Go to www.quickforms.com to learn more, or contact us with questions at support@quikforms.com.

Episode Transcript:

Intro: 00:02

Welcome to The Customer Wins podcast, where business leaders discuss their secrets and techniques for helping their customers succeed and, in turn, grow their business.

Richard Walker: 00:16

Hi, I'm Rich Walker, the host of The Customer Wins, where I talk to business leaders about how they help their customers win and how their focus on customer experience leads to growth. Today is a special episode in my series on new and emerging solutions, and today's guest is Chirag Gandhi of Mili. Past guests in this series have included Mark Ovaska of Precept, Richart Ruddie of Captain Compliance, and Thomas Clawson of Slant. And today's episode is brought to you by Quik!, the leader in enterprise forms processing. When your business relies upon processing forms, don't waste your team's valuable time manually reviewing the forms.

Instead, get Quik! Using Quik!, you'll be able to generate completed forms and get back. Clean, context-rich data that reduces manual reviews to only one out of 1000 submissions. Visit quickforms.com to get started. Now, before I fully introduce today's guest, I want to give a big thank you to Rich Whalen of Equity Services Inc., who is also a prior guest on my show. So go check out his website at equityservices.com.

And check out Rich's episode on The Customer Wins. All right. Chirag Gandhi is the co-founder and CEO of Mili, a house of tailored AI agents for advice firms. He brings deep financial services experience from Deutsche Bank and Credit Suisse. Previously, he raised and managed 400 million at Kalaari Capital and Trifecta.

Mili has processed hundreds of thousands of advisor meetings, giving the team a front-row view of how to integrate AI into real advisor workflows. This is exactly what I want to talk about today. Chirag. Welcome to The Customer Wins.

Chirag Gandhi: 01:58

Thank you so much for having me today.

Richard Walker: 02:01

Yeah, man, you're doing some cool things. I think everybody wants to hear about it. For those who haven't heard my podcast before, I love to talk to business leaders about what they're doing to help their customers win, how they build and deliver a great customer experience, and the challenges to growing their own company. So I want to understand your company a lot better. How does your company help people?

Chirag Gandhi: 02:22

Absolutely. So rich at Mili, we are building AI agents that can streamline operational workflows for wealth advisors and essentially get them more time to be talking to their customers, serving their customers better, and take away everything that's admin that's mundane or that does not help grow the practice. So where we really started was with meetings and solving that workflow. Essentially helping you prepare for meetings, take notes, update systems of record, and like you pointed out in the introduction, as we've now done this a hundred thousand times over, naturally we've had discussions on where else can really, really help. And it's interesting.

I've just come out of a lot of quarterly business reviews that we do with our enterprise customers, and the impact that they are seeing in the platform beyond just meetings has been phenomenal. So I'm excited to share all that and more in our chat today.

Richard Walker: 03:27

Nice. I gotta admit, when I talk to customers, I mean forms is what I talk to them about. But the number one topic I'm hearing is the Agentic workforce. How do I they all have initiatives like how do we build agents? I've heard of people building it themselves.

I've heard of people buying things and trying to put it together. Of course, you know, jump and Xbox and these note takers took off and everybody was talking about them. Where do you guys see yourselves fitting in this world? Is this a build it yourself? Are you building it for them?

Is it turnkey? How does this get into market for a customer?

Chirag Gandhi: 04:02

Yeah, absolutely. So rich. We are built specifically for financial advisors. And when we say financial advisors, we mean anybody that touches the financial services life cycle of a client. So the wealth advisor is obviously a very key has a very key role in that journey.

But you also have a tax planner and estate planning attorney, in some cases an investment banker who gets looped in when it's when it's an M&A transaction. Right. And really serves all of these different cohorts. And oftentimes it's it's interesting when both of these both of these firms are using to serve the same end client. Right.

And the kind of goodness you can have with solving for workflows, not just internally, but collaborate better with other stakeholders, is is great for the for the customer. So that's what we are seeing. And when we started we had a had a very cautious approach to deploying safe, practical and compliant AI. And that has continued to be our ethos given the industry that we are building in. But for us, what's been eye opening is over the last 12 months, the conversation has really shifted from from how AI could actually be implemented in our industry to what all can we solve with AI.

Right. And, and, and we want to utilize it to the maximum potential. So that's been great to see.

Richard Walker: 05:33

Yeah. So I want to go back though does a customer get your product. And it just works. Do they have to configure it. Do they have to do any do custom code anything for anybody.

Because a lot of advisors have different workflows and different types of clients and and CRMs and all the tech stack.

Chirag Gandhi: 05:50

Yeah. So that was the first learning we had. Rich, which is you could have two advice forms separated by a few blocks in New York. And their need of a software like ours could mean very different things. So today, Mili comes configured out of the box with a set of 12 different meeting templates with integrations with pretty much the ecosystem across your CRM, financial planning, software custodians, so on and so forth.

But beyond that, where we where we really thrive as a team is, we love ensuring there is that customization for each form. And the way we do it is with something called the forward Deployed Engineer. So our Customer Success team and the engineering team steps in even before forms sign with Mili as a platform, and we ensure that all of their workflows meet the meeting types. What gets discussed in the meetings are all mapped on the mini platform, so it becomes their tool. It's not a generic software you've purchased off the shelf.

Richard Walker: 06:56

Yeah. So there is an implementation process, right? Your team is working with the customer to understand it and then deploy it in the way that makes the most sense for them. Does that mean you also get into custom prompting and custom coding to build these agents? Look, I run a software company, so I'm just really curious, like how how flexible is your architecture versus what do you have to build for customers?

And when they ask.

Chirag Gandhi: 07:22

Yeah. So and this is something I can only expect in a podcast with an entrepreneur like yourself. But so what we have Rich is we've built out the skeleton. Right. So no matter what the meeting is, you would have a set of fields you want to discuss.

There are a set of objects and fields you want to update in your CRM. And the customization then happens at the level of those elements, which is what are the fields that you that you really want in the meeting? What do you want to update in the CRM or your financial planning software? Right. And Mili is constantly forming.

Like, think of it like a client. 360 that's ever evolving. That's a lot of the metadata about the underlying client. And and that really ensures that no matter what that end outcome of a meeting is, Mili is there to deliver. Right.

Like it's it's about it's for us about building that dictionary out for the form. And once it's in place, it's it's locked and loaded.

Richard Walker: 08:22

Yeah. Look, I'm asking for a lot of different types of reasons. Part of it is you have the compliance facet, right? Nobody wants to let loose an AI within their systems if they don't know what it's going to do. And the fact that you're going in and helping it understand and configuring it to the needs of that customer, that that situation, that also helps to understand the compliance facet, which I want to talk more about.

But the second, second thing I think about is everybody's system is different. I mean, Salesforce, every Salesforce implementation is different. And to say it's cookie cutter and turnkey is a fallacy for for most software implementations. And I think a lot of people well, I don't know if this is totally true, but I think a lot of people say, oh, AI is a silver bullet. It'll just do anything and everything.

But there is a real world, like, you've got to mesh these things together, right?

Chirag Gandhi: 09:10

Yeah, yeah. No, absolutely. So there is initially for forms, we end up spending a good amount of time in making sure that we understand their syntax. And one example of that is we recently launched Milli office. And think of it like ChatGPT for financial advisors that talks to all their systems.

That's agentic in the way you you want it to behave. And if somebody says, who are all my clients in Oklahoma, Mili would have to ask. Okay, what address do you mean? In the CRM? Is it the delivery address, the primary address, the secondary address.

Right. Like and and these are fields that are at first not obvious to to Mili. Right. And it takes some amount of training to say this is what it likely means in the scenarios and and give clients the option right in the generative interface to say, why don't you pick one and I'll remember it for the next time? Bear with me for the first time here.

Richard Walker: 10:10

Yeah, now I like this, I like this. So going back to the compliance thing, how how do you make sure that your AI is not hallucinating? It's not using data in the wrong ways, using somebody else's data to report on something else. How do you silo this thing? What do you do?

Chirag Gandhi: 10:30

So, rich, if I was to go back, I had mentioned the journey has really been of two years for us, of which the first year for us has been perfecting the right design choices, right? And those will keep coming up in our conversation as it relates to the meeting agent. There are three key differences that we have that make this the most compliant implementation of of a meeting agent. Number one, we do not send AI bots to the meeting. So if I would have Mili on right now in the call we are having, Mili would not join as a third participant as an AI notetaker.

We feel the space between the advisor and the client is pristine. It means something and the client may joke and say, I don't know if I should talk about that right with AI in the room, but when the joke at the minute, right, there'll be some part. It could be a death, a divorce that they are just not comfortable talking about. Mili works silently in the background. We would still request for consent, but the consent is not a AI bot staring into your soul, right?

While. While the meeting is happening.

Richard Walker: 11:37

Into your soul. I love it.

Chirag Gandhi: 11:40

So. So that was number one. Incredibly hard to achieve. There are like five companies in the world that do it, including OpenAI notion. So we are in good company in that, right?

The the second has been no recordings. So Mili operates primarily through streaming. And the proof of the pudding is the notes don't come up after the meeting. They come up during the meeting. While the meeting is happening, you would see the notes come up.

Mili would keep structuring them, would keep correcting her own mistakes made in the past during the conversation as well. And that's delightful to see, right? That's equivalent to having an assistant take notes in the meeting for you while while the meeting was ongoing. And for us, that means we don't need to create a recording ever, right? Like we are able to take notes in real time.

And that was another technology unlock for us. And the third has been a very strong PII redaction. So even before anything goes to your amazing LMS, right? Like an open. I applaud Mili internally redacts anything like Social security numbers, account numbers, addresses and all of that is configurable.

So we work with enterprise partners that tell us what makes sense and what doesn't. And that makes sure that while we have a zero training policy with our underlying models, it's always good to be safe. So that's what we're able to achieve. So these three design choices have been important.

Richard Walker: 13:05

I want to know the magic of how you did that. I'm not going to ask. That's too long of a conversation. But man that that's quite the achievement to get those those things working together the way you did. That's amazing.

There's another aspect of agents. I mean, this has been a really popular concept for a year, year and a half. I mean, I know last summer, well, a year and a half ago, people started to talk about agents. You were onto this two years ago, it sounds like. Like, did you have this notion of AI should be agentic before people were really seeing it and doing it.

Chirag Gandhi: 13:39

Yeah. Rich. And like, while our, our website was all about meeting notes for the first year because we wanted to be very mindful of what we implement, our seed fund raised deck to our investors back then was what's our grand vision? Right? Where do we see all of this headed?

And and that's where we see AI being being deployed. Right. So it's all about what are all the moments in your day where you'd be like, man, I wish I could give somebody this context and somebody I could delegate this task to someone, right. And for a lot of these tasks, it's hard to delegate because these are small, small tasks, but all of them add up, and all of them sometimes require different pieces of the puzzle. For some, you may want to refer to your financial planning software for some of your trading software.

And it's it's all here, right? How do we get that context out in the real world for you to actually act on those workflows? Was always the vision where we're headed. And for me, it's personally meaningful because I feel the threshold of net worth for somebody to get a financial advisor in the US is pretty damn high. Right.

And I think if we can give advisors back their time, all of that goes back into saying yes to that teacher, right? Maybe saying yes to that veteran. So so that to me is the most fulfilling part of of what we do.

Richard Walker: 15:04

No, I love that. The concept of agent has taken a lot of different forms over the discussions I've heard. And, you know, you talk about like small little tasks that add up and it makes perfect sense that agent can go do that. Hey, drop the note into my CRM for me. I don't have to do that.

Schedule the next meeting. Possibly. I don't I don't know all the things that you're doing. How how big can an agent be like? Could it replace an actual person in a role?

Chirag Gandhi: 15:34

So the evolution that we are seeing, and I'll definitely answer your question as well on, on where does all of this go? But I think the building blocks for us are the three A's. With agents, you should be able to ask, automate and alert, right? And most of when people say agent workflows, most of them today refer to ask which is information retrieval. What was what was the 401 k value right for this household?

Or how many of my clients have I not spoken to in the last year? All of that is ask. Some of the ask can get really interesting, but it's more about getting the data in now. As you do this more and more as you have deep workflow understanding, which we have been grateful of our our customers to to share that with us, you can start automating these workflows and that need to be in a predefined path, right? Which is in an onboarding process.

We want to get these documents. Documents need to be scanned, uploaded. These are the forms that we need to fill. And this is how we onboard a customer right. So for for one of the customers we recently filled up 1400 forms as it relates to onboarding and migrating migrating their customers.

So so that's the second piece. But the third is alerts. And today this is where the human cognition comes into play right. Which is doing things before you are told to. And that takes a lot of understanding and context of the form, how you serve your clients or even taking a step back.

Who are your clients, right? And what are you doing for them? And that context takes time. Yeah. And with all these three pieces is where you can deliver a truly agentic solution for for the customers, which is what we are building towards as it comes to.

Can it do everything a human can? We believe anything that is just about applying the cognitive power to a job is what AI would eventually be able to take up. The pieces that will become even more relevant and special are the relationships that advisors have with their clients, and anything that requires physical dexterity, right? Where it's also about going to places and doing things which may not just be about the cognition. Those are really the two special edges that that humans continue to have.

Richard Walker: 18:09

Good. So AI is not going to take my clients golfing for me.

Chirag Gandhi: 18:12

Absolutely.

Richard Walker: 18:13

Yeah. Got it. Perfect. No, it's a it's a really interesting concept because there's so much fear I hear from people in the workforces. My wife's in healthcare.

I hear about in the healthcare world. AI is going to replace my job, AI is going to replace my job, and I don't fully buy into it because I think that I think that it's going to do tasks more than anything else. And if it can link a whole bunch of tasks together and perform a function, great. Maybe you just don't have that function as part of your role anymore and you don't worry about that. And that seems fair because who wants to do all the tasks?

I mean, really?

Chirag Gandhi: 18:53

Absolutely. I definitely feel the narrative here has gotten a lot ahead off of the ground data that we're also seeing in the economy, even beyond financial advisors. And and I mean, yeah, trust. Trust the leadership to overstate the job losses to AI, right. Versus saying we probably underdelivered on our quarterly results.

Richard Walker: 19:15

So yeah. Okay. So you started this conversation by saying you're exploring what else Mili can do, right? You're solving all these types of problems. What is it that you've found really cannot do or probably should not do it, other than the human like we just talked about being present and taking the hard calls of sorrow or whatever.

Like, what are you finding as insurmountable right now?

Chirag Gandhi: 19:42

So there are really, I would say, two different buckets of what Mili cannot or should not do. One is as it relates to compliance in the in the industry, right. So if you think about it, SEC has been has given a free hand to AI for most purposes as it still forms a point of view on AI. But one thing SEC is very clear on is AI not delivering investment advice, because imagine OpenAI asking everybody to invest in the same stock, right? That wouldn't end well for anyone.

So so controlling market narratives and driving market action is not something anybody wants to be doing right now. It's okay to be informed with AI while you take those decisions, right? The second thing that we are seeing is anything that that is client facing, it's we've always found it better to have human in the loop. And for an advisor to eventually sign off on the deliverable, on the financial plan, on the onboarding kit, or even a follow up email. Right.

That just says, hey, lovely, lovely meeting with you. And in our case, 80% of the emails on the platform actually go with very, very minor edits to the to the clients. So Mili does a great job. But that said, it's always advisor in the loop and clients would often come back and say, I still not walk to the car in the parking lot of the office right before I before I got the follow up email. So there is a certain timeline you should send the email out in.

But that's been that's been the that's been the most interesting. So tasks where you have a very high impact on the client or it's client facing, I would say is always better advice to to be reviewed before before anything happens.

Richard Walker: 21:35

No, that's that's totally fair. So here's another curious thing I have when I talk to customers and they say, oh yeah, we're investigating building our own agents. What advice would you give to people? I mean, I don't know how to build an agent. I've used agents, software development.

We have agents. And I just tell Claude, build an agent so it builds an agent. I don't know the right way to do it. I'm just curious. I mean, you spent your last two years doing this.

What are the pitfalls? What are people not understanding of what it takes to build an agent?

Chirag Gandhi: 22:06

So there are three things as we see it. Number one is just the amount of investment, be it dollar or be it time. And. today the best of engineering talent is is not given to you on a platter, right? And the best talent also works.

Wants to work at companies that are inspiring, that are again building for the future. Okay. I mean, for for them, working at a a multi-billion dollar area is not the most exciting proposition, right? Because they'll do it once, but with a technology firm, they can do it several times over. And the relevant dollar cost as it comes to actually delivering the services over a period of time.

And for a lot of firms, that cost is just prohibitive, or with the amount of time it will be in the roadmap for the next years to come, right before it actually gets to fruition. So number one is is really that number two is readiness from the standpoint of data. And today data exists across multiple different silos. So ensuring that you have the right data mapping you have the right data hierarchy on in case of a conflict. What do you refer to.

Right. That's very important. And at Mili, we have built our own proprietary engine to say, okay, for the different data sets. This is generally the puzzle, right? And we keep evolving that with more, more learnings in the industry.

And the third is and this is something that actually a client of ours told us and it has it has stuck with me since you should really look inward as a firm and say, am I solving this only for myself? And is is am I the only benefactor of an agent like this, or does it make sense to solve this for multiple advice firms? If it's the former, customize it, build it internally, make it your IP. If it's the latter, it's. You're always better off partnering with someone because AI, at the end of the day, learns from data, right?

Like so the more context you have, the more mistakes you make, the more the underlying system gets better in in a multiple Industry. Independent studies. Mili has ranked the highest when it comes to accuracy of meeting notes action items. Right. And if it if it's at 97%, the way I would look at it is it's 3% more to go right.

Because there is no bar for error in the industry we're in. So so that's the continuous improvement that you have to look at. And I often joke with my team about, it's become so easy to build right now that everybody should be an entrepreneur of some sort. Right. You should always be building right now it's you could write a few lines of code and have a system that works for you.

So really, it's all about what edge cases do you want to be solving, right? Like what edge cases are you most interested in in troubleshooting that will keep you up at nights. And yeah. And do what? And that could be something that you build for your clients for example over a weekend.

And that's great.

Richard Walker: 25:08

Pardon me. You've made a really, really good case for the build versus buy. That's been going on forever about technology. Like build it yourself. You have to maintain it.

You have to invest in it. You have to do all this work, buy it from somebody. It's their core expertise. You get 8,090% of what you need. You don't maybe get 100%.

And AI is introduced, this concept that everybody's talking through of like, well, it's so fast and easy to build. Should I just build it myself now? And I think there's a lot of pitfalls in there. And I'm also seeing a lot of people come around going, no, it's not as easy as you thought. There's a whole bunch of people like, oh, I went on lovable.

I built an app this weekend. You can't sell that app. It doesn't work the way you thought. It worked for you. Yeah, yeah.

And, you know, there's another interesting facet to this, which is right now there's not everything exists that you want. And I'll give you a case in point. This summer, I built a solution that builds products in Amazon for me. And I mean like enterprise industrial grade systems workflows and, you know, databases, all these things for software companies. And it worked really, really well.

And I have this framework and it's still rudimentary. I think somebody's going to build something similar. It'll be a commercial product. It'll be like, oh, here's my idea, go build it for me, right. But that doesn't exist today.

So then I'm in this quandary of, well, do I bring that product to market? Because I built it and it's working and it's I think it's a hard thing because a lot of firms are inventing things as they go. But a firm like yours, where you're fully dedicated to this problem, you know, you're going to solve it. So let me ask if you've solved another type of problem. We have all seen the bias that llms have towards us.

You're absolutely right. That is the best question to be asking right now. You're right on. Let me tell you why. Right.

Those types of positive affirmative responses. In fact, I heard today somebody saying that he and his friend were debating on who's the best soccer player in the world. and they each asked AI and they got their answer and they were like, see? Because it was biased. How are you?

As you guys evolve, how are you protecting against bias? And not even necessarily hallucination, but just this weird bias that AI is bringing right now.

Chirag Gandhi: 27:27

So rich. So internally we have multiple facets by which we measure AI systems. But AI discrimination is really one where we have like now a 20 pager or a 30 pager, just like policy document where we are stress testing our systems across a variety of different metrics with like predetermined, predetermined scores on where we want them to be. And all of this obviously happens on synthetic data sets. So like synthetic made up transcripts of conversations between advisors and clients and ensuring that that doesn't happen.

Right. And and that bias could be in multiple forms, right? It could be as much as just being positively affirming on on something. It could be the information is just not accurate or it's not comprehensive or it's there is no brevity. Right.

Like you could the analogy we would get is advisors would say before the tools they used, they would have word salads. They would have like thousand words pasted on the wall. And it's just not helpful.

Richard Walker: 28:35

Right.

Chirag Gandhi: 28:35

So, so there are a lot of different guardrails of which bias discrimination. All of these are one. And as we have done this, we've also realized there is no right answer, which is why today ChatGPT as a platform is also struggling. And and there for them. The answer has been personas where Rich may want to talk to an AI that's friendly, that's supportive.

Right? A young kid may want an AI that's just bashful. And and maybe they're into something specific, right. So how do you really solve for that? Good thing is, we solve for financial advisors that generally aren't that are fairly homogeneous versus the broader internet.

But even with that, we love giving them the option to customize for their own use case and write in a few prompts on top. The most interesting ones would be I always end my email with this signature, right? Or when I start an email like I want, I want to touch upon a few personal aspects of of the meeting that we had, and I want to make a note about that as well. So all of that is is possible with a few lines now. So that's what we offer.

Richard Walker: 29:43

Yeah. So one of the techniques I've been using over the past two months probably is there's a concept, I guess, in software development, maybe just in general of Red team, blue Team, where the blue team owns the product and the idea and the red team debates it and pokes holes in it. And I didn't know this concept until about two weeks ago, by the way, and my CTO fully explained it to me, but I had this practice of building a panel of experts around a topic. So HR experts, legal experts, whatever. And I find that when you have that panel of experts and you get a collaborative conversation going, I'm seeing less of that.

Hey, Rich, you're right type of bias. And now I'm seeing more of different perspectives and views by having this kind of these different personas built into that single conversation I'm having. Do you guys ever do that?

Chirag Gandhi: 30:30

Got it. That's actually that's actually very interesting. And that's not something I have actually done in both for like I am also an avid AI user myself, but definitely tool of.

Richard Walker: 30:44

Choice for yourself.

Chirag Gandhi: 30:46

I would say Claude. And yeah, I find Google Gemini very left brained, very creative, artistic. And then OpenAI is the smartest model out there. But Claude gets both, right? Right.

Like, it's it's personable, but at the same time, It's so love the interface.

Richard Walker: 31:04

Yeah. One of the challenges I've had with ChatGPT. OpenAI is that it'll infer things or skip details. It'll start writing things for me and say, oh, the rest is a similar pattern. Just do the rest.

Claude doesn't do that. Claude's like, here it is, boom boom, boom boom. It'll just go and go and go and go and go and it'll give you all the detail and all the depth. They definitely have different personalities, if you will. And they've both been really, really good.

I have not played with many others. I've played with perplexity. I haven't really gotten into Gemini, so I didn't know it was so creative. Kind of left brain stuff. That's that's fascinating.

And now I you know, you're building Mili for the advisor. Do you guys use Mili internally as well? Does it serve your purposes in some way?

Chirag Gandhi: 31:49

It absolutely does. So like we obviously have our own meeting meeting templates, be it our customer reviews, feedback conversations, the onboarding conversations a quarterly business review. And it's collaborative in nature as well. Right. So it's like just one person in a in a meeting really has to have million.

And everybody benefits from that. We only got the HubSpot integration live a few months back, and that was great for the team because internally we also use HubSpot. So so getting all of that data back in the CRM has been wonderful. So it's now that we're getting to see the same magic that our customers get to see from the product.

Richard Walker: 32:32

Do you think Mili will transcend financial services? I mean, sales or sales, right. Having conversations and tracking information and performing work.

Chirag Gandhi: 32:41

Yeah. So when we started Rich, that was that was the million dollar question, which is what do we want to be when we grow up? And I think unanimously within the team we landed on. And I've personally been very excited to be solving for this industry. This is an industry where I have not yet had sort of gray hair, but I'm I'm getting there.

Right. So like have spent a decade here, really understand the pain points. And with AI, like you rightly pointed out, as like building a system, building something 0 to 1 is easy. It's all about how do you impact a few forms very, very deeply, meaningfully and transform how they actually serve their clients. And we think all of that comes from depth, not breadth, right.

Like it comes from really understanding. And if you think about it, we are not the first note taker. That's not a net new insight, right? Like we've had hundreds of note takers preceding Mili. But the adoption of these generic note takers in our industry before when we started was less than half a percent.

Right? So when you're trying to be something for everybody, like it works for the ease of use and the shallow use case to get started, but it's hard to have that context. You're not talking to the CRM, you're not talking to the things that matter, and you're always limited in the potential. So. So that's what grounds us to this industry.

But I would love to learn from you as well. On how did you decide on on that for Quik! and how has the journey been?

Richard Walker: 34:17

Yeah. You know, it's it's a it's an actual question that we struggled with internally as well. So we're in financial services because I left technology to become a financial advisor. And I had to fill out forms. And so that was my knowledge, expertise.

That was the problem. I understood and I didn't try to build a product. I just built a solution for me. And six months went by with people saying, give me your product, rich, I want it. And I said, no, no, no, no, no.

Finally I said, okay, I guess we have something here. We should build it. But the truth is, forms are everywhere. They're in the government, they're in healthcare, they're in logistics, they're in everywhere. So forms is a universal problem.

And we we years ago had to answer the question of do we want to stay deep in financial services or wealth management, or do we want to go horizontal across all industries at some point? And frankly, I think I've always had it in my heart to go horizontal because data informs or data informs, right? It doesn't really matter. So that that's why we've partnered with people who are really deep in new account opening processes or workflows and integrations. Not to say we don't have some, but our goal is really just to enable forms, period.

Get your data to and from the form, get them signed, get them processed, make it work within any workflow you want. And what's fascinating because you say dive deep. We chose to dive deep on the topic of forms, not necessarily the wealth management industry, which we know well. But now we're moving beyond that into banking and insurance. We have immigration law.

A bunch of lawyers use our product because it's forms at the end of the day, right. But there's so much depth in what we do that nobody else is doing it. I really think that Quik! is unique worldwide. We're waiting to see maybe there's somebody else doing it the way we do it, but I think, I think we have a lot of runway in front of us. I'm realizing we're coming up on time or maybe past it.

So I better get to my last question. But before I get to my last question, what is the best way for people to find and connect with you? Chirag.

Chirag Gandhi: 36:17

So we are at get AI. We've recently got a fresh paint of code on the website, so please do check it out and give us any thoughts feedback and my LinkedIn is is open love conversations on on LinkedIn. And some of my best sort of conversations have happened on that platform. And yeah go by Chirag Gandhi. The second name is easier.

So you'd get that spelling right? Yeah.

Richard Walker: 36:47

Well, you're in the show notes so people can come to find that as well. All right. Here's my last question. Honestly, my favorite question who's had the biggest impact on your leadership style and how you approach your role today?

Chirag Gandhi: 37:02

So rich. For me, like. Idolizing leaders has like as a concept has definitely evolved with me over time and over time. I've learned to really pick up different skills from different leaders on on what I adore. And at the end of the day, humans are also imperfect, so what serves them well in certain conditions doesn't serve them well in other conditions, right?

So it's all about picking what you align with with the most. And I think in terms of the leadership style, one person that has continued to sort of I've continued to learn from idolize has been Jeff Bezos and the kind of culture that he has built at Amazon with a lot of attention to detail. on the smallest things, right from how things get stacked in a warehouse, to obviously the broader sort of strategy. Right. But getting getting each of these pieces right, with a lot of intricate design choices and everything boiling down to we are all about the decisions we take every day, right?

And everything that we do as leaders, as team members. So and all of that then determines what our clients value us for, right? The trust, the transparency or the belief that if if it's Amazon, I'll get my order right. So that's the kind of culture that we are also building here at, at Milly. And something I think is, is like enduring right.

Like and is really timeless.

Richard Walker: 38:41

Yeah. No. I have huge admiration for Jeff Bezos and Amazon. One of my favorite companies, for so many different reasons, especially because of my doorstep getting packages on time or early. Yeah.

Man, this has just been an awesome conversation. I hate to say goodbye at this point, but I want to give a huge thank you to Chirag Gandhi, co-founder and CEO of Mili, for being on this episode of The Customer Wins. Go check out his website at getmili.ai. That's getmili.ai and don't forget to check out Quik! at quickforms.com where we make processing forms easier. I hope you've enjoyed this discussion.

We'll click the like button, share this with someone and subscribe to our channels for future episodes of The Customer Wins. Chirag, thank you so much for joining me today.

Chirag Gandhi: 39:26

Thank you so much, Rich.

Outro: 39:28

Thanks for listening to The Customer Wins podcast. We'll see you again next time and be sure to click subscribe to get future episodes.