[Emerging Tech] How To Transform Enterprise Data Into Insights With Suvrat Bansal
- Quik! News Team
- 2 hours ago
- 32 min read
Suvrat Bansal is the Founder and CEO of Clarista, a platform that connects fragmented enterprise data and delivers real-time, AI-powered insights for faster decision-making. He is a seasoned data leader with over 25 years of experience transforming data into business growth across global financial institutions. Before founding Clarista, Suvrat held senior roles, including Chief Data Officer at UBS, as well as leadership positions at Credit Suisse and Morgan Stanley. He is known for building scalable data strategies and pioneering approaches that simplify complex data environments into actionable intelligence.
Here’s a glimpse of what you’ll learn:
[02:39] Suvrat Bansal shares an overview of Clarista and solving fragmented enterprise data challenges
[04:27] Managing data strategy at trillion-dollar scale and impact on user experience
[09:02] Security-first approach and handling enterprise data across global regulations
[11:34] Streaming data versus copying data and eliminating traditional data models
[15:31] Suvrat talks about Clarista as a real-time data processor and flexible deployment options
[17:52] Problems with legacy data architectures and unnecessary data duplication
[20:05] Why Clarista is data-first, not AI-first, despite heavy AI usage
[24:44] How designing products at the pace enterprises can realistically adopt
[29:49] Blending AI with compliance, governance, and enterprise constraints
In this episode…
Organizations struggle to make sense of fragmented data spread across multiple systems, slowing decision-making and increasing operational costs. Traditional approaches rely on copying and restructuring data, which creates inefficiencies and risks around accuracy and security. With the rapid rise of AI, many leaders are left wondering, how can they unlock real-time insights without adding more complexity?
Suvrat Bansal, a data and AI expert with decades of experience in global financial institutions, explains that the key is to rethink how data is accessed and used rather than how it is stored. He emphasizes a data-first mindset, where organizations connect and process data in real time instead of duplicating it across systems. Suvrat highlights the value of simplifying data access through business-friendly structures, automating workflows with AI, and aligning innovation with enterprise constraints like security and compliance. He also advises leaders to focus on outcomes and user experience, while gradually introducing change at a pace organizations can realistically adopt.
In this episode of The Customer Wins podcast, Richard Walker interviews Suvrat Bansal, Founder and CEO of Clarista, about transforming enterprise data with AI. Suvrat discusses real-time data processing, eliminating data duplication, and balancing innovation with compliance.
Resources Mentioned in this episode
"[Emerging Tech] The Future of Financial Advice Automation With Eden Ovadia" on The Customer Wins
"[Emerging Tech] Turning Online Reviews Into Growth With George Swetlitz" on The Customer Wins
"[Emerging Tech] Compounding Institutional Knowledge for Growth With Ian Karnell" on The Customer Wins
"Revolutionizing Fintech Sales Strategies With Wim Van Lerberghe" on The Customer Wins
Quotable Moments:
“Our fundamental objective is really to drive organization scale. They have the right people, they get stuck in the wrong jobs.”
“We don’t copy any customer data because our belief was you can do all data processes on the fly.”
“People don’t want to believe this thing, right? Just because you talk about it, doesn’t mean that it’s believable.”
“We will always consider Clarista as a data-first company, because without that, we kind of feel there’s nothing.”
“Software is not confidence. What you should be driving on is here is the outcome I’m seeking guys.”
Action Steps:
Focus on real-time data access instead of duplication: Shifting from copying data to accessing it live reduces inefficiencies and improves accuracy while enabling faster insights and minimizing risks from inconsistent data.
Simplify data for business users: Presenting data in intuitive, business-friendly formats helps non-technical teams engage more effectively and make better decisions without relying heavily on technical support.
Align innovation with enterprise constraints: Introducing new technologies gradually ensures compatibility with compliance, security, and operational frameworks while increasing adoption and reducing resistance.
Prioritize outcomes over technology: Keeping the focus on desired results helps teams stay aligned and prevents unnecessary complexity by ensuring technology serves clear business goals.
Use AI to augment workflows, not replace them entirely: Automating repetitive tasks boosts productivity while maintaining oversight and control, allowing teams to focus on higher-value work.
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.
Powered by Rise25 Podcast Production Company
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. Some of my past guests have included Eden Ovadia of FINNY, George Swetlitz of RightResponse AI, and Ian Karnell of VastAdvisor. And today is a special episode of my series on new and emerging solutions. And today's guest is Suvrat Bansal, founder and CEO of Clarista. 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 Quick! 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 Tom Waisnor of Advintro for introducing me to Suvrat.
Go check out their website at advintro.com as they specialize in outsourced augmented B2B business development and sales resources. I also had their founder on my show, so go look for Wim Van Lerberghe. I always say his name wrong Van Lerbergheto to hear about how they helped their customers win. All right, Suvrat Bansal is a seasoned data leader with over 30 years of experience turning fragmented enterprise data into scalable, trusted intelligence. As former global chief data officer of UBS Asset Management, he led the data strategy supporting over $1 trillion in assets.
Wow. I get to talk to you, man, today. He's the founder and CEO of Clarista, an AI powered, intelligent platform that bridges fragmented wealth technology into real time, decision ready insights for advisors and enterprises. So welcome to The Customer Wins.
Suvrat Bansal: 02:15
Thanks for having me, Rich. Super excited to be here.
Richard Walker: 02:18
Man, I'm so excited to talk to you. This is gonna be fun. So for those who haven't heard my podcast, I just love to talk to business leaders about what they're doing to help their customers win, how they built and deliver a great customer experience and the challenges of growing their own company, and especially in this age of AI. So Suvrat, let's understand your business a lot better. How does your company help people?
Suvrat Bansal: 02:39
Yeah, look, there's no surprise. Of course, you know, financial services are all about information economies. You know, everything from people on the frontline looking for information about customers or trying to acquire new customers, analyzing markets. Everything is digital. We know that.
But that digital means systems, platforms, third parties. It's all over the map when you're trying to construct an answer, right? It's just too much. It's too costly, too heavy, too many teams involved. What does is just make that whole process happen in a very lighter way, right?
It's a unique technology which can bring information from different places to produce insights, which you're looking for augmented with AI. Of course, nowadays, like everyone will do that for productivity gain, but it's much deeper, right? In terms of connecting your enterprise data beyond your personal chat AI, right. What the context you have in personal chat is limited, as we all know. Secondly, the same thing transpires in operational workflows.
I mean, similar to what Rich was saying about Quik! forms, essentially what's happening with the combination of engineering and AI, you are in a very strong position to automate most of your workflows today and turn human intervention and manual input into just a review process. And that gives you scale. And I think our fundamental objective is really to drive organization scale. They have the right people, they get stuck in wrong jobs, and we want to eliminate those wrong jobs and give them time back to scale that growth.
Richard Walker: 04:01
Oh, man, look, there's a bunch of stuff in there that I want to talk about. I'm still stuck on this trillion dollar number, to be honest. Yeah. You know, because it's funny. I mean, you're a data guy and I would expect you to say, oh yeah, I managed, I don't know, 20TB or petabytes of data or whatever.
Yeah. But nobody gets what that means when you, when you translate it to money and the importance and value. Yeah. What did that look like really to you?
Suvrat Bansal: 04:27
Yeah. Look, let's start back with the experience part, Rich, because finally it's all about whose experience are you going to impact by doing any role like data. Now remember the chief data officer. Unfortunately they are the only people with data in their title, although every person in the organization is dealing with data. So you have a unique responsibility.
And I think, you know, for people who take it seriously, know how heavy that job can be. But essentially it's really about understanding the experience layer of information at every function and then executing it in a way which is not a bunch of end user computing. Right. I mean, you can throw Excel sheets at some problems. You can throw a database at some problem.
But when you think about an organization and feel the gravity of that role given to you, you have the responsibility to balance the horizontal experiences of thousands of people and the vertical complexity of, you know, hundreds of years old organization with, you know, thousands of systems and end user computing and the regulatory frameworks and auditing frameworks. And that's what my specialization was. I started in tech. I always believed in technology being like a personal thing. First, try it before you deploy it.
Every new technology I brought in, I tested it on my own on weekends and took coursework where I didn't understand the technology well enough, and then I was very good at scaling it. I was again fortunate that I never walked into an existing job, so I had to create things very rapidly without compromising with any sustainability of what I was creating in the long term. I think having mastered that, you know, like those things is what finally the, you know, kind of lightens the weight on those shoulders of those roles. Like you need to have that experience and capabilities and frankly, enough failures to learn from and how to avoid that. And then the biggest benefit in the data data world has been the constant evolution of capabilities.
So you're not stuck with the only way a 20 year old can do things. Now that's more of a process and blueprint. People still follow, but a lot of it is about educating myself first and then then educating my stakeholders on what's possible. People don't want to believe this, right? This, because you talk about it, doesn't mean that it's believable.
So you have to make it believable. You're starting with the experience layer. So I think, yeah, it's a trillion dollars, but it transpires into the experience of every function client, be it client facing teams, product facing teams, operations, regulators, compliance, and then it transfers at the end of the day into an organizational view of how you're innovating it. And we used to do that, like, of course, we had 18 streams crossing different stakeholders at the organization level. We had, you know, of course, I mean, just to give you a scale of this, we were adding 1.5 million holdings a day.
Wow. A day. Right. Like, so, you know, these are different scales at which people are operating in these organizations across, as I said, hundreds of systems and people. And, and they all, every person is busy doing their work.
Right. So someone who's trying to figure out how to optimize that, I mean, of course there is prioritization and budgets and all that goes with the territory. There is no magic here, right? You work with the right layers to build the case. You get the right sponsors at the table.
You, you know, go to the leadership and tell them how you're going to execute one thing which can benefit 15 things. But that's the rule, right? Yeah.
Richard Walker: 07:53
Look, I, I love that you started by framing it as experience because every single person in business cares about data one way or another, whether they know they should or not. It's a different question, right? But we measure ourselves against our data. We look for data, we look for patterns. I was just doing pattern recognition this morning because the clients' usage was off the charts.
And it turns out they're actually using our system every 27 seconds, then every five seconds later, and then every 34 seconds later, and then it repeats that same pattern. Yeah. What are they doing? There must be some monitoring, right? Data matters.
And that experience that you give to your constituents, your, your customers, your internal team, etc. all that matters and what you're talking about is how this turns into experience. So now let's go to Clarista because I think what you clearly stated was that your product is meant to improve experience. Yeah. So one of my curious curiosities, now I'm a data person, by the way. Yeah.
Quik! is really a data company when it comes down to it. How are you? How are you using AI to surface the data and turn it into intelligence? And how secure is this, you know?
Suvrat Bansal: 09:02
Yeah. So, you know, that's again, my experience coming from like, look, banks like Morgan, Credit Suisse, UBS, right? You hear these names and you put an immediate belief that these organizations must be doing something right in their security posture. Otherwise you would hear it. Right?
Right. They're brand big enough to know that they have to do it right. So we got trained very early on, like I think most people don't know, right? Like it's a very strict organization that has a separate data residency law. You have to read it.
When I joined my job at Credit Suisse, I read, read all of that. I equipped myself with that. And it's a very strict security profile on what data can leave the geographical boundaries of Switzerland. That doesn't mean where you're logging into a system from. Everything is recognized.
You're traveling. You're in the UK right now. Nope. You can't log into your system. So, you know, when you train yourself with that kind of discipline, access controls encryption of course.
Right. But we're talking about on the fly in a digital interface where you will be blocked because you're not, you've left that geographical boundary from geolocation perspective, and your IP is no longer coming from Switzerland. These are the kind of strictest things we have lived through. Right? Yeah.
And we still need to create that experience. So it doesn't matter what the regulatory guidelines are. And I won't call them constraints because everything is solvable. These are architecting things, right, to finally deliver on that experience because teams are global, right? You're not creating a Swiss team and a US team for everything.
It's not it's it's subscale. It's not optimal. So I think yeah, just training on that. So essentially what we did in Austria, we remove the inertia out of the process. So why are we so secure?
We don't copy any customer data because our belief was you can do all data processes, every data process on the fly. You can view data, analyze data, join data, visualize data, and you know, go deep run AI without copying it in any vendor platform. The challenge was people thought you have to create this rigid model. A lot of tech companies still operate like that. Like I have a data model.
I'll bring information to my data model and then give you a login to that.
Richard Walker: 11:17
Yeah.
Suvrat Bansal: 11:17
And our view was why, you know, when Netflix, you're not downloading Netflix movies on your iPhone anymore. Are you?
Richard Walker: 11:25
Streaming it?
Suvrat Bansal: 11:26
You're just streaming it. It's caching enough on a browser and it kills it when you sign off, that browser is gone, right?
Richard Walker: 11:34
So.
Suvrat Bansal: 11:34
So when you take that, and that was my first when I left, I said just the thesis. Is there a personal level? Why are we moving all this data? It's not as heavy as a video stream. Of course, there are some complications.
There's one stream and packets and optimize that versus, oh, I have to kind of connect to multiple systems and formulate an answer to what people are looking for. But that's what we made possible in Clarista without compromising with the thesis. And I start with, I don't have your data. If you don't have an account, we'll create it.
Richard Walker: 12:08
This is fascinating. I hadn't thought about this. You're right. But how? How?
Yeah. I mean, data lives in data stores and databases and, you know, object stores. How are you streaming it?
Suvrat Bansal: 12:21
Hearing and seeing it is just getting confused in the marketing language. So take snowflakes as a company. Snowflake came out and said, we will have real time shares. You can just connect data into different systems. Like we all know snowflake usage, a storage layer of cloud S3 or ADLs, right?
So essentially what they have is a technical schema, which, behind the scenes, can connect to any physical store sitting in the cloud, right? But they limit it to their store because their whole revenue model is you have to live and breathe in a snowflake ecosystem so that we can charge by computer, right? Right. Nothing wrong with that. But from a client perspective, there are things fundamentally missing in it.
One that means I can only use that technology of on the fly storage versus compute separation when I use snowflake. Why it's a technology, use it anywhere. So I took that approach and said, I will give you a business schema. It's not limited to engineers. Think about data products.
We talked about that. Data views. Just think about spreadsheets like keeping it very simple. You log into a spreadsheet. There are ten tabs.
Each tab has something useful in it, right? The headers are what are meaningful. The data speaks to the header, right. So the headers are a business schema. The second tab contextually should be related to the first step.
Otherwise, why are these tabs in the same spreadsheet? Third tab. Each tab can reference another tab through calculations. Now you just take the spreadsheet and say, okay, let's turn it into an enterprise solution.
Richard Walker: 13:56
Wow.
Suvrat Bansal: 13:56
So you have headers that will fill real time, including any real time computation from wherever the data sits can be Oracle on prem, SQL on prem cloud Data Lake Cloud Warehouse and snowflake and Databricks and Salesforce and workday doesn't matter anymore, right? Because that's just the network. All we did was say the networks are there. We are using it on iPhones. Let's turn it into an enterprise solution.
So that's part one. Problem solved. Right? Rich like, and by tying the business semantics to the technical source, why I give this example of spreadsheet headers is why normal business people use spreadsheets so much because they don't like the complexity of technical data. They don't understand what a schema is.
They don't understand what a table is. I mean, try finding the right term for it. They don't care. Yeah, it is my view.
Richard Walker: 14:43
Right.
Suvrat Bansal: 14:44
So construct these views on the fly. If you're not copying the data, you're not limited to the rigidity of creating physical views of data. So you have completely federated it. The federation engines have been around for a long time, but by connecting it to this semantic layer, you're just quadrupling the benefit of it. And once you connect that schema in Clarista, everything in Clarista runs on those headers.
So I can run SQL on it, visualization on it, I can talk to AI on it, any context I need. I just need to add that context to those headers, right? I don't care where the data is anymore.
Richard Walker: 15:22
Wow. Wow. I look, I don't know if this is normal information, but this feels like a major breakthrough. Maybe this is really how you guys.
Suvrat Bansal: 15:31
The biggest, I would say, is that we came from very complex enterprises and we thought the world was there.
Richard Walker: 15:41
Yeah.
Suvrat Bansal: 15:42
And we tried to solve very complex problem thinking. We're solving it for the world. Since then, we have only backtracked Rich, because the world doesn't even know that these things are solvable. Forget that. They have realized the problem and they're ready for a solution.
Right. So and then. And remember where we started. Who cares? Give me the experience.
Richard Walker: 16:05
Okay. I have another curious question for you. How do your customers see you from a processing standpoint? Are you a processor now of their data?
Suvrat Bansal: 16:13
We are purely a real time processor of their data. And we are very flexible because we run SaaS. So customers who are lighter, like say, just we have banks, we have credit funds, they're using us for SaaS. Then there are certain clients who are like, hey, look, I want to internalize into my own cloud networks, technically everything is possible. You know, it's just you're compromising with the, with the speed of, you know, improvements rolls out if you're running it locally, which we know.
Right. Typical challenge. So we're very flexible. Technically that's not a problem. Any cloud, any platform we can run on it.
But yeah, that was the liberation that, you know, we're working with. We got into discussions through, well, two insurance companies only to realize they have a massive problem of legacy data and cloud not talking to each other. They, you know, and they're like, first, can you do AI matching and merging of this data? I said done, right. Like here it is like, and yeah, one and a half day, I think we went back here it is.
It's a very old problem for them. And then they say now that we've merged it, can you create a virtual warehouse for us. Here it is. Right. That's the technology.
Like the answer should not be. Let me go back to the drawing board and architect this. It should be. Here it is. Yeah.
And if you can solve problems, Why I went into that is because these fundamental layers were the inertia of a rigid data model, ETL pipelines running everywhere. There is a place for everything, but it needs to be highly selective based on purpose versus the blueprint of that's where you start this whole medallion architecture. Like I get in and I mean, I am a vendor, I'll listen to any client company. I was a client, and I will basically tell the vendor to leave the room if they talk about medallion architectures.
Richard Walker: 17:51
Right?
Suvrat Bansal: 17:52
So now as a vendor, it's completely fine. Like, look, if that's what you want to do, but let me share my view. It's copying data, 90% of the data in three locations.
Richard Walker: 17:59
Yeah.
Suvrat Bansal: 18:00
Why?
Richard Walker: 18:01
Why, right. You lose your source of truth and why.
Suvrat Bansal: 18:04
Oh, well, because it comes as raw. I need to transform it. Why not transform it on the fly at consumption time? I don't know if that's possible. Oh no.
It's just not even possible. Like, I don't know if it's possible. So I think a lot of it immediately gets into I don't know that's possible.
Richard Walker: 18:18
That's what I'm saying. Like this is a breakthrough. Like you have figured out something most people don't understand, including me. I didn't realize this. And the.
Suvrat Bansal: 18:25
The reason I'm citing.
Richard Walker: 18:26
It is a heavy process that's expensive and, you know, slow. Yeah. But you're saying it's not it's not.
Suvrat Bansal: 18:32
I mean, we run it every day. The clients are running like we have a client so demanding that they create these very complex credit funds, very large funds they use to sign around 68,000, you know, distribution and capital call notices a month through Clarista. So Clarista is a very fungible framework, which you can extend beyond insights for all the workflow optimizations. But coming back, like they said, oh, I have these, you know, five tabs and each tab has like around eight charts. And I need five second performance guys.
And it's all real time. But the fact is we made it happen, right? So this is possible. I think our clients push us enough to make it a possibility wherever, you know, they feel the experience can be improved. But I owe it all to the evolution of networks and the technologies which we were seeing when we started.
Clarista Snowflake had not become that big. It became big in front of our eyes. So we were like, awesome. Now people will understand the power of separation of compute and storage, But this is just talk. Rich people don't understand these concepts.
They just care about the experience.
Richard Walker: 19:40
Well, again, it goes back to your first premise. People care about their experience, not the data. They need the data they don't understand how to, how to use it and how to get it. They just want it delivered. Yeah.
Okay. I want to ask another question. And look, I have a fundamental question, but I really want to get into the depth of this. Do you consider Clarista an AI first company? And where are all the places that you're using AI in your company?
Suvrat Bansal: 20:05
Yeah. So this is a philosophical debate. We will always consider Clarista as a data first company, because without that, we kind of feel there's nothing to do. Yeah. So we don't jump to let's talk about AI.
First, start with, let's talk about your data first, right? I have to source it. Finally, I have to link it like not even if I'm not copying it, I have to rely on it. So let's just make sure it's ready because otherwise you'll tell me that the AI answer is not wrong.
So how do I validate that the data is right before I say the answer is not right? Right. So and then yes. Everything we do from a client workload perspective is AI driven because we kind of feel like we look in the digital space, a long time coming, if we have turned these things into digital anyway, like it's not physical movement of cards anymore for trades, right? We have seen those ones in the late 90s that are still kind of floating around somewhere, but like people don't even know they were punch punch cards and all that, right?
So there is nothing physical left other than servers. So I think, yeah, in that it's just, it's, it's, it's due, it's just that, you know, there was, I don't know what was missing. But yeah, I mean, with AI, what I feel is the biggest benefit is that people don't have to imagine the experience. So they like, just like you were talking about, I think we should talk about it like your whole autonomous developer using AI, right? Some people are extreme where you're going rich by saying, I can create autonomous development with AI agents, but in the whole spectrum now you can see people from asking questions to, hey, look, can you read this from my calendar?
Or can you go further and now create a CRM system for me? People are taking different legs, but the fact that's happening in all like what happened last year, and that's the exciting part. That's the exciting part. I'm not saying the LMS came one year back, but I'm just saying it became a lot more public conversation one year back. And I think with that, what happens is when you get to the clients, they already know the roadblocks they're hitting, where are those roadblocks going to be?
Enterprise data security controls and auditability and regulations, which, you know, you don't have to worry about in your personal AI and those takes take us immediately back to how secure the solution is and how you handle data. So we are an AI first company in the experience layer, both operational workflows, automation and the insights, automation. but the conversation very rapidly goes towards where the clients get stuck in their personal experience, which is all about the enterprise considerations. Yeah, right. So hence we are always going to be the data and security first company.
Richard Walker: 22:42
No, I love that. I love that you keep your focus because your core value is to focus on the data. Yeah. And even though I call quickly, you know, we're, we're driven, we're AI first. Now we have been for a few years.
It's still about the forms and the data. Fundamentally, when I say I first for my company, I want my team to use AI to make their jobs easier, better, faster. I want to use technology to go do things we couldn't do before. Yeah. Yesterday for my team, I laid out the next three to five year roadmap for our product, which is only possible because of AI.
It's been a dream for 20 years. And now it's possible because of AI.
Suvrat Bansal: 23:19
Oh, Rich. I mean, and, and the biggest challenge you'll have is that 3 to 5 years will turn into 3 to 5 months, and you'll be back at the drawing board.
Richard Walker: 23:26
Oh, I hope so. Oh, I hope so. Oh my gosh. Because this is a product roadmap, not a sales roadmap. And the product is going to unlock more and more market opportunities for us.
That's what I'm excited about. Yeah. Because you know, I think about it, we're limited by forms limited by what forms we have. So what if we could get more and more of those faster and faster, better and better? We can do more for our customers.
So, you know, I'm actually in the midst of building something to create software that builds software. Autonomous software development is a pipe dream. It's on the next to impossible list. And I'm actually achieving it. I'm just this is what I spend all my fun time with.
My kids are like, what are you doing now? But one of the things I'm doing to build a web application for this, this autonomous system is it's an AI first application. So when you log in, it's, you've got your dashboard and you can go to administration and add users, but you can also just go to the AI and say, add a user, Show me my status. Start a project. So I'm thinking how do we build web applications to be AI first?
So the experience can just be I'm going to talk to this site. I don't have to click on anything. Just take me here. Show me this. Do this for me.
Are you guys doing that or are you ahead of me on this? I hope so.
Suvrat Bansal: 24:44
So. So I would not, no no we're not. I mean, at the end of the day, Rich, just like the smart people like you in the world. Right. So for me to sit here and claim we are ahead of anything, it's kind of ignorant, right?
We're doing what we're doing, right. So the way we approach this is when you are doing that autonomous software development, and you're ready to roll that out and you go to an enterprise, they will, they will. That autonomous software will be very data hungry and audit hungry right at every step of the way. They will ask you to like, oh, well, here, I need this to be triggered for my version control. And I hear I need this review to be done by my architect.
So what they will do in the early days is take you back a little bit from the vision to where they are, and try to find a middle ground by injecting their current teams and current process into a completely autonomous process. Yeah. This is the ground reality of enterprises. They move slowly, right? Yeah.
And because they have taken years and decades to institute those controls, every innovation initially gets pulled back to comply with those controls. Next comes how you automate the controls. Right. So we all know the journey. So I can fly very fast today on AI because we already know the pullbacks.
We have to methodically make sure that that step gets automated. So hey you data quality should turn into agent right. People should.
Richard Walker: 26:20
So I'm hearing something here. I'm hearing that you are predicting how enterprises will react to your product, so you're rolling out features at their speed, not your speed.
Suvrat Bansal: 26:30
Yes, that is exactly what you're hearing because we feel that's where we add unique value.
Richard Walker: 26:36
Yeah.
Suvrat Bansal: 26:37
Right.
Richard Walker: 26:37
Smart. No, it's smart because this is what I learned with forms. Do you know how many vendors I've seen come into the forms marketplace and say, we're going to revolutionize it? I know, totally change it.
Suvrat Bansal: 26:47
I hear you.
Richard Walker: 26:47
People don't want that big of a change. They're like, just make my forms electronic. Yes. Just make them fillable. Now make them signable.
Don't give me the whole thing, you know. Yes. Don't plug into my head. Make me think. And it shows up.
That's too much. Yeah, yeah. You're right.
Suvrat Bansal: 27:03
And also, like I, I have this compliance person, I have this CSA, I have this like, where will they unplug into your process? You may say. Yeah, they don't need to. They'll say, no, they need to. Yeah.
So I think it's recognizing those realities. There's nothing wrong with it. It's like, look, I'm coming from organizations where I come from, controls take a lot of time to issue a policy. It used to be a year-long process. Yeah.
Because once it's issued, you have to audit it. So people used to first challenge me back on why this policy. One more. Right? It's not like, oh, write a policy for data governance.
There's a lot of that noise I used to hear. I go in and write a policy. I say, yeah, what about the business risk officers push back on you on having to monitor one more policy and which employees will need to be trained on it, whose training schedule needs to include it? How will you audit it? Who will monitor the results?
How will it be fed back to the quarterly attestation of the business risk.
Richard Walker: 28:00
You're giving me nightmares of SoC two. We went from 80 to 180 controls in four years. Like more. More. No, we need more.
Suvrat Bansal: 28:09
Yes, yes.
So, you know, I think and so part of it is just exactly what you said, so we are SoC two, ISO 27,001. We have to be GDPR because we have clients in the UK. We, you know, we're working in a tangential space of healthcare. So you're also, you know, our HIPAA certified. So I think there are cases where, and, you know, this tangential space, the way it's evolving, like give you a perfect example, insurance company, a client of ours came to us and said, look, anytime the advisor is selling this insurance product.
So we were focused on wealth. The advisor has to work with the client to manually fill this massive insurance form, all the health details and all this. Can you please help us automate that? Yeah, right. Very similar to your case, right.
But what we are seeing is these boundaries are merging with AI too, right? Traditionally, there'll be handovers from one business entity to another business entity. And now those boundaries are merging without compromising with those controls. So I think, you know, it's just about recognizing those early on so that you can build your architecture accordingly to be flexible. Some clients will fly, some clients will take a methodical approach, right.
But you need to be ready for both.
Richard Walker: 29:19
Yeah. All right. So look, in the last year, you mentioned the past year how much things have changed. A year ago that was the year of agents. Now everything's going to be an agent.
Nobody could agree on what the definition of agent is. Yes. I think to this day we can't agree on what the definition of agent means.
Suvrat Bansal: 29:35
Yes.
Richard Walker: 29:36
There's probably 12 flavors on a spectrum of what an agent is. So I'm curious, did you buy into that and name that an agent in your system? Do you guys claim to have agents working on behalf of users? Does it make sense to even call them agents in your world?
Suvrat Bansal: 29:49
Look.
The way we think of it, we kind of instead of an agent, we kind of think about it as like, and this is where the dialogue is started. An intern or an analyst, they are a new entrant in your organization. They can do certain things very well because they've been educated well, you have hired them. They have enough knowledge. They don't understand your organizational considerations.
Richard Walker: 30:13
Yeah.
Suvrat Bansal: 30:14
That's how we think of AI in general. So we call them AI analysts. You can call them agents. Look, I care so little about this. Whatever. I don't play an intellectual debate on any of this.
Let's call it for whatever. So for insights, we tell people that, look, when an analyst is gathering, you want to do analysis on a new offering. As a family office, someone has come in in a venture capital fund and they're offering a new fund and you want to analyze it. What happens? You analyze it by itself in the data room of 50 documents because, you know, the venture fund has just sent you those 50 documents.
Then you analyze it against the existing investments with that fund. So if you're invested in another four vehicles, you'll say, look, how is this doing? And by the way, some unique things which I love are in highlighting a person's concentration risk. The same person who was successful in two other funds was highlighted in the third from a marketing perspective, without highlighting capacity constraints. Great question to ask.
Right. But these things get missed out because you're so focused on financials versus things which matter. Then you have to analyze that against 140, let's say, other investments you have made with various fund managers. Personally, I will do it. Good job.
50 documents is not huge. You can upload it and analyze it, but everything else it can't, right? And there lies the rub. Like you, you want to be able to. So would we call that an agent who's thinking incrementally at different levels of depth with the common purpose which has been given, which nowadays are called skills.
You can define a skill like section one, data room, section two, existing investments. Section three. My entire book.
Richard Walker: 31:50
Yeah.
Suvrat Bansal: 31:51
That's easy to explain. The how part is the hard part. And now if I solve that, you can call it an age I. At that point I care less. Have I solved it and is it the experience the client wanted?
Get it, take it right. Like, is it an agent? Great. It's an agent.
Richard Walker: 32:10
Bring it up because of the things. You're the point you're making. Like you really shouldn't care about the how. You should care about the outcome. Yet we think of agents because we're humans.
We want to identify some role, function, and skill. If we call it an agent and we give it a name or persona, it feels normal to us. Oh, that's an engineer. That is an analyst. That's a security expert, right?
So from a marketing standpoint, it sometimes feels better to tell people, hey, we have this collaboration of agents. My autonomous software has over 70 unique agents working together. Yeah. It's crazy, but do you really care? Do you really care how they work?
No. You care that it works.
Suvrat Bansal: 32:48
And I think.
Rich in enterprise space, it can actually concern people.
Because by the sound of it, it feels like something is happening behind the scenes, which I don't understand.
Richard Walker: 33:00
Yeah.
Suvrat Bansal: 33:04
So I think I would highly encourage people like us who operate in the enterprise space, like you and I do, to explain this as it is, don't worry about the language, right. I think because people spend so much time and we have client conversations, like every vendor is coming to me on AI trying to pitch it. Same thing in different ways. And I said, yeah, just have a conversation. We'll get to a product at some point if you want to, but let's just draw on the 25 years of corporate experience.
We can chat anytime. People love that. They want to have conversations.
Richard Walker: 33:35
Oh yeah. Thankfully, I met you before we had this conversation that we're recording and I love talking to you because it's just, it's easy.
Suvrat Bansal: 33:43
It's fun. You're interested. Great. Yeah.
Richard Walker: 33:45
You know? Yeah. Look, I have to wrap up this.
Suvrat Bansal: 33:49
Of course.
Richard Walker: 33:49
And I hate this because I want to keep talking to you, but we're going to talk more before I get to my very last question. What is the best way for people to find and connect with you?
Suvrat Bansal: 33:58
Yeah. I mean, look, the easiest is of course, you know, have a quick look at our website, freshair.io. And at least it tells you basics on, you know, what we're doing and some examples, some case studies, some client testimonials just like basics, right? But to the extent you're already trying testing different AI tools. I'm pretty sure that within the enterprise you want to hear what are those enterprise constraints and how do we overcome them?
So you can reach out to me on LinkedIn. Just search for me. I was fortunately, unfortunately given the unique name, right? I mean, not too many to find. So you just search for it and you'll find me and you can reach out on LinkedIn as well.
So yeah, it's very easy nowadays to find people and reach out to them. I'm always open for dialogues. Discussions. Doesn't matter. Yeah.
Just all I encourage people is when it carries a stigma of like, oh, I don't want to have a sales conversation, but you don't have a sales conversation with people coming from the corporate side, you know, people kind of actually are thinking a lot more on your behalf than you would. And you don't feel alone in that, right? Don't just you don't have to go to five conferences to learn. That is my point, right? Yeah.
If nothing will at least tell you everything, you should be asking how you should be like, you know, how, how we have thought about it. And that's just not if nothing becomes your checklist is how I look at it. Some clients have actually asked me, can you put that together? I'm like, okay, whatever. It's fine.
If that's what helps you, that's fine too. I just start with, what are the experiences and the outcomes you're trying to drive? Is it more operational efficiency? Is it more in size? Is it both?
Okay, great. Where do you want to start? And then how have you thought about your data in wealth? Unfortunately, it's still very separate. And not every vendor is a very open API ecosystem, as you know, Rich, you worked in that space.
So yeah, they are just focusing on that. How do you want to make it work? Like people, I tell you this, this is like exactly where a client dialogue versus reality comes in. So clients will have a dialogue. Do you have a built-in connector?
Now I try to explain to them why connectors were needed. If you had a fixed data model and you had to build something which is coming in into a fixed form on the right. We have no right side.
Richard Walker: 36:10
Yeah, it's.
Suvrat Bansal: 36:10
Just a head scratcher for them. So like, what does it mean? I say, look, I can't control what Salesforce will deliver tomorrow as an API change. Do you want me to sell you connectors? We don't charge by connectors, modules, users, all that is as a buyer.
Think about it, Rich. You know you hate this.
Richard Walker: 36:27
Oh yeah.
Suvrat Bansal: 36:28
Yeah. Because you're trying to forecast everything at the time of contract. Who has the time?
Richard Walker: 36:32
Yeah, I know and then yes. How do you manage your long term costs too? Right.
Suvrat Bansal: 36:37
Yeah. Throw it all out right now. I think what you should be driving on is here is the outcome I'm seeking guys how fast and what cost. Right. Yeah.
And are you ready to commit to that? So why do we commit to outcomes in phase one? I tell people whether we are working through a partner or we are working directly with the client, because that's the only way to deliver confidence. Software is not confidence.
Richard Walker: 36:59
No. I love that you said this because I haven't heard anybody else articulate it. But it is true. People who have lived in corporate and then go start companies, they care about solving problems. Yeah, everybody wants to make money building a company.
But what we care about is solving problems. And that's where my head has been. Okay. I have to get to my last question. Wrap this up.
So total change of style here because this is my favorite question. Who has had the biggest impact on your leadership style and how you approach your role?
Suvrat Bansal: 37:29
Oh wow. So I will have to say three people. Rich. And I'll quickly tell you who and why. Caroline Arnold when I joined Morgan Stanley, I was there for 13 years, but she was the vice president of investment banking technology, and became a very senior MD over time.
Right. But what she taught me early on was how to take care of your people. My next boss was Paul Martin, head of operations for asset management globally in Morgan Stanley. And he taught me what integrity truly means. And how do you display it every day consistently?
And my third boss was Thomas Heinzel and UBS, whom I would like to mention. And he told me like he really taught me that innovation is not incremental. You have to go all the way. You have to learn everything. Like this guy will give me, like, we have a meeting tomorrow.
Read about quantum, whatever you can. Let's come to the meeting prepared. Right. So that's the kind of people and there are building blocks. I was very fortunate because what Caroline built in me was teams.
But Paul told me about how you operate. And then Thomas taught me about how you deliver.
Richard Walker: 38:44
Nice, I love it, I think.
Suvrat Bansal: 38:46
So I was very fortunate to have these people in my life to be able to like, you know, mention their names and what they contributed to my life.
Richard Walker: 38:53
I agree, I agree. I, I'm in a way, I'm envious that I haven't had a boss for 24 years. I didn't get these experiences in the same way you did. I love what I'm doing so I'm not going to complain. Nothing too.
Suvrat Bansal: 39:05
Complaint. Yeah yeah yeah. No, you're doing great. You're self-taught, right? I think I could have used those influences at that time of my career, for sure.
Richard Walker: 39:13
Yeah. All right. I want to give a huge thank you to Suvrat Bansal, founder and CEO of Clarista, for being on this episode of The Customer Wins. Go check out our website at clarista.io. And don't forget to check out Quik! At quickforms.com where we make processing forms easier.
I hope you enjoyed this discussion as much as I have, and we'll click the like button, share this with someone, and subscribe to our channels for future episodes of The Customer Wins. Suvrat, thank you so much for joining me today.
Suvrat Bansal: 39:42
Thank you so much. So much fun. And of course, our conversations will continue. Thank you.
Outro: 39:48
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.
Comments