Insightful Analytics: Empowering Action in the Digital Workspace | Interview with Zalak Trivedi, Product Manager at Sigma Computing

Insightful Analytics: Empowering Action in the Digital Workspace | Interview with Zalak Trivedi

===

Ryan Purvis: Hello, and welcome to the Digital Workspace Works podcast. I'm Ryan Purvis, your host, supported by our producer, Heather Bicknell. In this series, you'll hear stories and opinions from experts in the field, stories from the front lines, the problems they face and how they solve them, the areas they're focused on from technology, people, and processes to the approaches they took that will help you to get to the scripts for the Digital Workspace inner workings.

So, Zalak, welcome to the Digital Workspace Works podcast.

Do you want to introduce yourself, please?

Zalak Trivedi: Of course. Yeah. well, my name is Zalak and, I've been a product manager for about eight years now and I have a broad experience across multiple industries and build products, in social media, payments, customer support, and now business intelligence at Sigma.

For the past few years, at Sigma, I've been leading the product direction and vision for the embedded analytics, which is really all about companies delivering insights or analytics, to the end users, typically their customers, but could be any external stakeholders like vendors or suppliers as well.

Something I am really excited about, is how our embedded analytics product can go beyond just analytics and help companies build entire data applications. So this is something as a company we are marching towards as well in FY 25.

Ryan Purvis: Great. And then, as I said, the standard question is, what does the digital workspace mean to you?

Zalak Trivedi: Yeah, so when I think about digital workspace, I think of three main things. The first is collaboration, second is personalisation, and last is, access. Like unified access, but really governed unified access comes to mind. It means really that everyone has the right tools and resources to do their job, and be successful without sacrificing any security, In the process, right?

And what I've seen is this helps really companies from off any scale from startups all the way to enterprise really scale it in a way faster, because if you have the access to the right tools and right resources, you're always not dependent on someone else to do your job. And this is something that we strongly believe at Sigma as well, which is data access for everybody and everyone in our company,

doesn't matter which line of business you are part of is a data analyst. Everybody uses Sigma on a day to day basis. And they have access to the right data that they need. And we really plan on taking the collaboration even to the next level. Right. So Sigma was originally built for the business users and the spreadsheet users.

And there are about 2 billion of spreadsheet users in the world. Uh, and which was, which was a great start. Right. But, as we kind of uncover and go into more and more deeper into the enterprise space in the BI lens. We also encounter SQL and Python users. and we are making sure that doesn't matter what your skill set is,

you are able to collaborate on a single workspace or workbook. If you want to use spreadsheets or SQL, if you want to write Python, and then latest with our, AI capabilities. We just want to write English. We'll help you create these amazing dashboards in your workbooks. So really, I feel like the future of work is a digital workspace where every part of business is using data and collaborating effectively, because really, you shouldn't be making decisions without looking at the data.

So you just empower all of them.

Ryan Purvis: So couple of questions as you speak, and I liked your sort of three things you broke it up into. So much so that I was so busy writing them down, it's fine. I mean, so, so would you, I mean, is Databricks something that's similar to you guys, or do you use Databricks as part of your solution?

Because what you mentioned in the sort of unified experience of whether it was Spark or Python or whatever, you could use that in Databricks using Kafta and the rest of it to bring data together and build out with, I think Python as well at that point, to work with your data, but it never felt to me like an end user tool, like you really needed to know what you were doing.

Zalak Trivedi: Yeah. So we don't, it's not really exactly similar to Databricks, right? So all these cloud data warehouses like Snowflake, Databricks, or, you know, Redshift, Postgres, BigQuery, it doesn't matter what data warehouse you have. We, like at Sigma, we are a very lightweight layer on top of these warehouses. So we don't really store any data, but we help you analyse the data.

So we, connect to all of these data warehouses and we are a really pretty front end that sits on top of the back end, which is the data warehouse. You can then analyse the data and really write SQL, but all of the queries are being then passed on to the warehouse. Don't see them a lot, as, someone similar to Sigma.

We see them as partners that, we work together with.

Ryan Purvis: Yeah. And that's fine. I mean, I think in this world, don't have to compete with everybody. but if you can find that complimentary thing. That's why I say, I mean, my experience with Databricks back, I mean, talking probably five years ago, a lot has happened in that space.

Zalak Trivedi: Yeah.

Ryan Purvis: It was very much a data scientists tool, right. And, and it was really UI on top of and Spark, which were. You know, Apache tools that you really also needed to understand Linux and all the rest of it to use them. But I like what you're saying because it's very similar to one of my other businesses that I'm involved in, which is Finxone. Where we are UI layer on top of all the APIs for FinTech apps.

Zalak Trivedi: Right.

Ryan Purvis: And it's designed and very similar stuff. And even when you said to the point of writing English to build the app, we get on the same route exactly. So it's a very interesting, synergy, I think. and then I also liked your mention around the sort of accessibility, security, all that kind of stuff, because that, plays a part very well in my information security background.

So I like that too. So what is the biggest thing that your customers struggle with? What's their biggest pain point? Frustration.

Zalak Trivedi: I think as In terms of, so I, oversee the embedded analytics product. So let me speak to that a little bit. So embedded analytics really means you are embedding Sigma and providing analytics to your customers, right?

So it could be in two different phases where. One could be, you already have a product and you want to augment your product with analytics. And the second could be, you have a new idea or a new data product that you want to launch, and you want to use Sigma, as your front end again, to provide those insights to end users.

So I think the biggest thing that people really struggle with is, the product itself, right? It has to be built for. Several different personas. So when every customer is really different, every customer is serving a different persona because it could be a store manager, it could be an analyst. It could be someone on in a restaurant who's looking at the restaurant insights.

So I think the thing that people struggle with is how do we create or curate like the right experience for their users. And I think this is a common phenomenon across the entire BI industry and specifically in embedded analytics, because you really are building for a persona that is a, like a tertiary customer to you.

and then our customer need to figure out what's the right experience based on the tools that we provide. So that is the, I would say the biggest struggle, but what we do at Sigma is we make that extremely simple. With the technology that we build, we allow our customers to curate their experience from all the way to just view only dashboards where I don't want all the bells and whistles.

I just want my end users to look at the data that I'm providing them to highly curated where I want to just provide access to the data, and I want my end users to do all the analytics themselves, kind of pushing the self service all the way to the end of the user. So we just make it really simple through allowing our customers to curate that experience.

But typically, that's what we saw as a big challenge early on when we embarked on this journey of embedded analytics.

Ryan Purvis: Yeah. And it's funny because I'm busy solving a problem. In fact, I just signed up for a trial. so we'll be talking again after this, because one of the challenges with wrangling data, especially in Excel, once you go past a certain number of rows, especially using pivot tables and that sort of stuff, Excel becomes a bit of a problem child.

So I can, I can definitely see the value proposition for what I'm doing right now. Hence I've signed up because You know, compute that you need sometimes and even a bit of insight, a bit of intelligence to the data. So if you see a whole bunch of rows and you can see common things, let's say it's, like a project with a bunch of tasks you want to interpret those tasks around, you know, whether they complete the date they started, the date they ended, they overdue?

Are they, on time, the resources that are assigned, all that kind of stuff. They're common patterns to looking at that data. Yeah. Yeah. And I wonder if you guys do anything like that, where you're giving some insight to data that you see often because you're seeing it often.

Zalak Trivedi: Yeah. So I think just to kind of back up a little bit, right?

So overall, we see that Sigman customers , when they're looking for a tool like Sigma are running into what you just mentioned, right? The scale is a big issue. With Sigma, you can, like you can have 2 billion or over 2 billion rows and you know, it loads the tables in seconds. The other challenge that we've facing just overall at the Sigma, customer who's coming to us, they feel that as soon as they look at the dashboard, the insight really dies there.

They're not able to take any action into it. And what we do is we allow exploration, like self serve explorations. Where if I, as a product manager, let's say I received a dashboard from my analyst, the first thing I think about was, well, how can I get this thing a little bit more granular or how can I drill down into this one specific feature that I've launched?

I want to look at which users are using this and understand the metrics around the usability or usage. So like I think the first question is okay, how do I drill down and Sigma solves really well for that? And again with embedded analytics, you can extend it to your customers So it's like all the power of Sigma is in your own product if you want.

Ryan Purvis: Yeah

Zalak Trivedi: And I think, kind of going back to the question where, because we are seeing this data, so frequently, are we providing any insights? So I think that's where, our new AI initiatives would come in, where, because Sigma's, like, one of the core principles is we do not store your data itself, but we do have access to your metadata, right? So column descriptions, table descriptions, who's the user who is, creating these workbooks? Is it, Jake from Sigma who's like a super awesome analyst, or is it like Zalak, a product manager who knows what he's doing, but he's not like a decision maker for the entire company, right?

So we have all that metadata, available. And this is in our plans where we will be incorporating all of this information into providing you with recommendations on, what workbooks are important, what dashboards are important, what kind of data or data models that you've created in Sigma are important.

And let's say if you join the marketing team like Sam, you will present it with all the important information, that typically a marketing person would look at when you log into Sigma. So again, very personalised, which is what I was talking about earlier. Like personalisation is very important.

Me as a product manager would really like all the, you know, feature flags and usage metrics, but Sam would like, okay, how many, Website hits we are getting, or how many, hits we're getting on this podcast that we are doing. So like more around the, communication and customer marketing.

Ryan Purvis: You're not storing the data itself. So you're like, so if I gave you a CSV file or an Excel spreadsheet to upload, you wouldn't hold that data for reporting.

Zalak Trivedi: No.

Ryan Purvis: So would I have to put that in like a one drive or I'm just trying to think of for the most simplest user who doesn't even access their own database. They just pulled out their sales pipeline, for example, and they want to do some nice reporting. How would you solve that problem? Or was that just one of those where, you can't because you don't store in the data.

Zalak Trivedi: Yeah. I think we heavily rely on, the data warehouses. In the future we may have like a program where we allow like a Sigma generator like Snowflake account or Databricks account Sigma own Snowflake account. But I think at this point we are very much focused on serving the customers who have these warehouses So if you let's say have a csv, we will just put it in your warehouse and then just analyse the data as we work.

Ryan Purvis: So you're not, so you're not targeting the one person band, you're looking for the company that's got their thing. And that makes sense.

Zalak Trivedi: Yeah. And they also help help these companies, right? So let's say if you are a company who's still figuring out your data strategy, we connect you to these warehouse based on your use cases.

And we have really good partnership with no Snowflake Databricks and so on. So we work with you and figure out what your data strategy is and then help you be successful.

Ryan Purvis: Yeah. Yeah. I mean, I suppose the other thing is you could just put it into a, I mean, an Azure data store or an S3 bucket or something like that and just suck it out of there, I guess.

And then tell me about, I mean, would you consider you guys a sort of low code, no code platform in some respects?

Zalak Trivedi: I mean, at this point, we've been pretty much no code internal to Sigma, but as we are involving like SQL and Python, which will be probably low code is what we are targeting, because if you are a large organisation and you have a very small team that is focused on, data science, we don't want you to go and buy another product, right?

You can, you should be able to do what you want to do within the same product that the rest of your organisation is using. And we'll provide you with the right tools to do that. And then you can collaborate again. Collaboration becomes super simple if you are all in the same product, even the same dashboard per se, right?

Ryan Purvis: Yeah.

Zalak Trivedi: So now we are kind of moving towards a low code, mechanism, but historically we've been just a drag and drop, you know, change things in the GUI and make your dashboard look pretty.

Ryan Purvis: Which is what everyone wants in the end. Because that's, that's what takes the longest for most companies is to spend hours and hours moving a pixel from one side to the other side, because, someone wants a different, way.

Zalak Trivedi: And what we found. Kind of taking this to the embedded analytics. We are also very similar. We have a very lightweight, like iframe based deployment. We've made a choice where we don't want you to code each and every, element. We just provided the controls in the UI for you to able to configure it.

And we have several like parameters that you can use. So your actual deployment is pretty simple. All the work you're doing is just creating this workbook and deploying it with the right row level security to your end user. So I see my data and you see our data. So we kind of have that philosophy where, I mean, yes, everybody can write SQL and you can code if you want, but why?

Right. If you can make it super simple, just do it through the GUI.

Ryan Purvis: Yeah. And then, I mean, tell me a bit more about your sort of product management style. How do you do things? I mean, what gives us a better insight to that?

Zalak Trivedi: Yeah. So as I said, I've worked in several different industry. And every time I go into a new company, my first three months is just understanding and becoming a product expert before I actually make a lot of changes or contributions or try to create a roadmap for the company.

So my product style essentially is to. Step one, become an expert in the area that you are going to be leading and then start talking to the people who have been here before you. So, at Sigma, I interviewed several of our SEs because they are really very close to the customer and our prospects, and they are also, Really technical enough for me to get all the information from them.

I talked to several prospects, as well as part of like, I just, I was just, sitting on the calls and understanding all the problems that they had to make sure that I'm understanding the persona and the space, itself. And then I think, the second part of my conversation was with our engineering lead.

So, at Sigma, we had engineering pods and then PM sort of signed to different engineering pods. So I sat with engineering manager and understood like, okay, what's the, how's the way the team is operating before? Again, as a new person, you don't want to go in guns blazing. You want to make you slowly assimilate yourself into the whole process.

So sit with them, understand how the sprints are going, how, like what's the mechanism that they are following today, and then start making small changes as you see fit. Rather than just going, okay, this is all wrong. Let's do it this way because I've done this before. That's not the way I approach things.

And once you have your, a nice foothold in the company where no people understand what you do and people respect your opinions, I think that's when, you start pitching ideas. You make sure that, everybody, all the stakeholders that you have are aligned on anything that you want to do.

And then just not take these projects forward. So at Sigma, we have quarterly planning. So before the quarter starts, we have all of our projects, and pitches ready, and we have alignment all the way to leadership, and then we actually just start, developing quarter over quarter. Let me just bring the customers along on the journey through private betas, public betas.

And, I work very heavily with our product marketing team to make sure we have the right assets like blogs, documentation before everything is launched to the customers.

Ryan Purvis: And how do you, I mean, in all those people you're talking to, you're going to have SEs, they're going to have sales engineers for those don't know what SE is.

That's the, that's normally the technical person in the sales process. And then you talk to whoever else, how would you rationalise what they're asking for? Cause sometimes it's a want, not a need. and then the other thing is how do you rationalise all these things when there's potentially conflicting priorities?

Zalak Trivedi: That is such a great question. The want versus need thing comes up so frequently, right? Because a lot of times, these requests are emotionally charged because someone is trying to close the deal and they really want something to be done. And that's the job of a product manager to isolate the needs from the wants.

And the way to do it is someone is pitching a solution rather than giving me a problem to solve. That is clearly, you know, I just need this to close this deal rather than, you know, this is going to be a long term benefit for the company. So at Sigma, all the product managers, do a really good job of kind of isolating this and digging down into the real problem before we actually start pitching solutions.

And I think you just get this from experience over time where you know who the right people are to talk to to understand and then you bring in your engineering manager or engineering partner to understand if like is this really something that's possible or is this something which is a pipe dream and we would probably never do. So we're just being very transparent about those things as they come through the pipeline and in terms of prioritisation like early in my career I used to use these frameworks but I think as you kind of go in and get more experience like these frameworks are really good starting points. But then as depending on your company and your team you can evolve based on that. So I always look at what's our company's most critical priority for the year and does this new thing that has been brought to my plate actually aligned with the goals that Our company leadership or product leadership have set. And then based on that we make a decision on prioritisation. Like for example, as I said this year we are heavily focused on bringing data applications to market.

So how does this new request that came in, h elped me move closer to that goal. Right. There will always be exceptions, right? If there's a large enough customer and who is going to be heavily impacted by this one small thing we do you always rerationalise but that has to be an exception not a rule. By rule you always follow a high level strategy that you've set because you set it for a reason, and if you always get distracted, you are never going to like do what you said you were going to do.

Ryan Purvis: But now I mean, I want to unpack it a little bit more because that's kind of the general answer, but now you've got you've got a whole bunch of stuff that's been given to you.

Do you have way of measuring priority to have like a framework that you use yourself says this is going to high business value, for example, the technical complexities low or medium or high.

Zalak Trivedi: You can use rice framework. I think again, I think that's like the foundational framework that most people use.

But again, it depends very much on what are your priorities for the quarter again?

I've never used rice actually.

Zalak Trivedi: Yeah, I think that was probably the first one that I used.

Ryan Purvis: okay. I'm having a quick look at that.

Zalak Trivedi: Yeah, it's like reach impact confidence and effort. Okay. Yeah, I think all frameworks are kind of like similar and that's, out of where. How many customers you're going to reach?

What is the impact on the business? How confident you are that this is going to be successful and what's like your dev effort look like? And then based on that, you can do a scoring and, and decide what you want to do. But again, I think this is a good starting framework. You always evolve it based on your current circumstance and your company and your vision and your priorities.

Ryan Purvis: Yeah, no, I think, I mean, you've got to have a level of, evolution to whatever you're doing. Exactly. and also I think when you have experience, there'll be some stuff that you could almost get, get instinct on exactly to a framework. Okay. And then how do you get feedback from stuff that you've had built?

Do you go out and see customers and ask them how they find it or what's your process?

Zalak Trivedi: I think it's, two ways, right? So, when we are launching a new feature, we go through the process of private beta, public beta, GA. And in the private beta, I pick the customers that are willing to give us the most amount of feedback, because the first iteration is never going to be the perfect one.

Because as soon as we launch, there will be some new things that we discovered that we missed during the requirements gathering or when you start using it, you will find new things. So find like five customers for internal use cases, five for embedded analytics and identify if we have missed anything.

We go into the public beta after and expand it to all of our customers who want access to and then GA. So that's me asking for feedback. But then there, we also have two other things, which is roadmap reviews. So as the renewals come, people typically would want a roadmap review of what's coming up. And we do a split where it's 30 minutes of roadmap review.

And then we ask questions to the customer about the features that we've launched and they've used, based on the user analytics that we have. And if they're not using something, we ask them, is it the discovery problem? Is it that they don't have a use case for it and then kind of uncover more insights.

And the second piece here, which is again, after the release is, user meetups. We have a very strong community page now, which is growing. And we also then have user meetups based on the communities. So we have separate sessions again, where a lot of people either come to a New York office or San Francisco office and, we ask them questions there as well.

Ryan Purvis: And your background? I mean, you have a very interesting surname. Where is that ?

Zalak Trivedi: I'm from India. It's from the western part of India, Gujarat.

Ryan Purvis: Oh, I don't know where that is.

Zalak Trivedi: yeah, close to Mumbai.

Ryan Purvis: Okay, nice. And how long have you been in the US for?

Zalak Trivedi: I've been here for about 10 years now.

I came here 2014 for my MBA. Got my MBA from UC Davis, Bay Area. And then I've been in the Bay for, the rest of the time. Yeah.

Ryan Purvis: Wow. Okay, cool. Anything else you want to chat about?

Zalak Trivedi: Want people to kind of reach out to us if you, if they are interested in creating data applications, which is again, a really big, push for our company where, if you as a product manager or an engineering leader are looking to create applications, where actions are very closer to insights.

I think Sigma is going to be a great product for you and we want to learn from you and we want to build things that actually help you. So reach out to us, reach out to me on LinkedIn or go to www.sigmacomputing.com/product/embedded-analytics.

Ryan Purvis: Yeah, actually, you just mentioned something that I actually wanted to explore a little bit now.

So if you embed the analytics in an application, I mean, the old days you were doing it through iframe, it was really a one way integration. You were just showing the window to something else, but are you seeing now where, for example, I'm drilling into the data, I could actually drive an action from that. Okay, tell me what you're thinking.

Zalak Trivedi: Yeah, I think that's the biggest piece, right, in the olden days, where you look at this iframe, you look at this dashboard and that's it. You had nothing else that you could do beyond that. If it's embedded in an application, you see this data and then you go back to the application to take some action because you find something interesting.

What we are doing this year is we are bringing that action closer to insight. Where, let's say if you're looking at a list of tasks. And we will provide you with an action button where you can say, okay, I want to complete this task and it goes and updates your application as well. So think of like a mini workflow engine that you can create where if you click on a button.

An action can happen. An action can happen in your application. An action can happen in some other Sigma dashboard an action can happen in any third party system, right? So if you have a list of customers that you're looking at through the iframe, and say, okay I want to put them in my Next campaign that I want to run so you can select all of them and say okay send to HubSpot And then it'll send to HubSpot and put it in an appropriate campaign.

So we're trying to make it Very interactive, insights, rather than just looking at an inside and that's it. So and you can also create like record views, detailed views, and have chained actions, and so on. there's a lot we are unpacking here.

Ryan Purvis: No, it's fair. I'm very excited for that because that's exactly the use case.

I was thinking about where I'd like to try to product and see if I could get it to work within hours because I don't want us to go and build a whole new white tool to build dashboards. but I want something that we could use our data obviously to show the stuff and then have it interact with our system.

Zalak Trivedi: Yeah.

Ryan Purvis: So, no, that's exciting. Great. so LinkedIn is the best way to get ahold of you, Do you want to give that out so people have it?

Zalak Trivedi: Yeah. So reach out to me on LinkedIn. I think that's the best way to reach, but you can also just go to Sigma's website, which is Sigmacomputing.com slash embedded analytics, which is where you can sign up for a free trial for embeds.

Ryan Purvis: Fantastic. Great. That's been great chatting with you.

Zalak Trivedi: Awesome. Thanks Ryan.

Ryan Purvis: Super. Thanks a lot. All the best.

Bye.

Thank you for listening to today's episode. Heather Bicknell is our producer editor. Thank you, Heather, for your hard work on this episode, please subscribe to the series and rate us on iTunes at the Google Play Store.

Follow us on Twitter at the Dww Podcast. The show notes and transcripts will be available on the website, www.digitalworkspace.works. Please also visit our website, www.digitalworkspace.works. And lastly, if you found this episode useful, please share with your friends or colleagues.