The Exchanges

Q Yeah, that's great. Let, let, let's get into this. Like what, what are the five pillars?

A Uh, yeah. So, you know, first is freshness. Um, so freshness is relating to, um, Uh, uh, the freshness of the data. So for example, you know, it talks about, uh, media companies, you know, you can probably think about e-commerce companies or even FinTech company that relies on thousands of data sources, um, you know, arriving, let's say two to three times a day. Um, how do you keep track and make sure that thousands of those data sources are actually arriving on time? There has to be some automatic way to do that, but that's sort of a common reason for why data would break. Um, so freshness is one. The second is volume. Um, So pretty straightforward. You know, you'd expect some sort of volume of data to arrive from that data source. Has it arrived or not? Um, the third is, uh, distribution and distribution sort of refers to at the field level. So let's say, um, there's a credit card field that is getting updated or social security, um, number field that gets updated and suddenly it has, um, uh, letters instead of numbers that would obviously be, um, something is, is incorrect. So you actually need tests for that at the, at the field level. Um, The fourth is schema. So actually schema changes are a big culprit for data downtime. Um, oftentimes there's engineers, um, or other team members actually making changes to the schema. Uh, maybe they're adding a table, changing a field, c…

AI assessment note: “first is freshness... second is volume... third is distribution... fourth is schema... fifth”

Q Uh, question from Jason. How do you think about observability and insights? With that semantic knowledge of the data, do you see limitations to looking at data without this additional information?

A I, I will just start by saying I don't think that that's possible to solve. I don't think that a machine can actually infer that we can infer something without knowing that business knowledge. It's not possible, and that's also not what we attempt or attempting to do at Monte Carlo. Um, I do believe that there is a certain level of automation that we can and should introduce that we have not introduced to date. Um, and that by introducing that level of automation, we can reduce our, our customer's team's work. Work from, you know, uh, 80% manual work to 20% manual work. So we can actually, you know, with the automation cover, cover 80% of reasons for why data downtime incidents happen and allow data teams to reserve their work for the top, you know, few percentage of, um, issues that only they will know about. Um, so we're not here to sort of replace data teams or to sort of know or kind of understand the business context. We don't, we don't attempt to do that. Um, We're only attempting to make, to make data teams lives easier. And so, uh, in today's world, most data teams actually spend a lot of time writing manual tests on, um, uh, things that can be automated.

AI assessment note: “I don't think that a machine can actually infer that without knowing that business knowledge.”

Q You know, do you have, like, any kind of, like, anecdotal story where, like, uh, having, having data that was wrong was, uh, not just annoying, but, like, led to, like, really serious consequences?

A Yeah, for sure, and happy to give some specific examples. You know, companies actually report, uh, numbers to the street, um, and accidentally report, report the wrong numbers, or about to report the wrong numbers. Um, that happens, um, more than you'd like to know, probably, Matt. Um, or, you know, for example, one of our customers is Fox. Um, Fox streams, um, you know, major events like the Super Bowl is an example, and as you can imagine, they're tracking, they're tracking lots of information about, Um, those events, right? Like how many users, where, where are users spending time on which content and which devices? Um, and so the integrity of that data is incredibly important because decisions are made in real time based on that data. Um, or another. Another example would be, uh, Vimeo, um, a great customers of ours, um, video platform, uh, streaming company. They have over two hundred million, uh, users in fact on, on their platform and they use data, um, and have used data throughout COVID-nineteen actually, um, uh, to identify sort of new revenue streams, um, also make real time decisions about the users. So for example, if there's a particular, um, user that actually needs more bandwidth at the moment, um, for example. And so if you don't have the right data, Um, at hand is actually very difficult to, to give the, the adequate or right experience that you'd like for, um…

AI assessment note: “happy to give some specific examples. You know, companies actually report, uh, numbers”

Q Thank you very much. While we're on the topic, a question from Chow Yun in the group, uh, does data observability mean different things or different applications for Different modes of data structured versus unstructured real-time versus historical, or does it cover everything?

A Yeah, I think in general, what we, our goal with, um, sort of the term data observability is to apply it to data everywhere. Um, and, and obviously it has different meanings and different types of data. Um, especially if you think about unstructured versus structured data, um, you know, we're also seeing more and more sort of streaming. Um, so definitely there is, you know, lots of different, um, changes that are happening in, in the data stack and in how folks, um, think about making sense of their data and taking action on it. Our belief is that you need to be able to trust your data wherever it is and what, and whatever type of data it is. You know, with most of our, um, uh, sort of companies that we work with and that we see, um, you know, we, we spent a lot of time on the data warehouse and NBI, um, kind of where we started, um, so we spent a lot of time there. Um, we're seeing more and more folks, um, Sort of move to obviously different technologies. And so, um, our thinking is that in order to, um, build strong data observability practices, it has to sort of include, um, concept that we called, um, end to end, uh, meaning including wherever your, wherever your data is all the way from ingestion to consumption. Uh, the nature of data is that it changes, that it flows. Pipelines are added every day by new team members. And so making sure that your data is accurate at only …

AI assessment note: “obviously it has different meanings and different types of data.”

Q How do you determine if this, uh, an issue or not an issue?

A Uh, we actually use machine learning for that. Uh, so we sort of infer, um, kind of what does a healthy baseline look like, um, and kind of make assumptions based on historical data. Um, so we use historical Um, uh, data points, um, collect those, um, infer, sort of project what the future should look like or might look like for you, um, and then use that to let you know when, um, something is off. Detection is incredibly important, but it's just sort of the tip of the spear, if you will. Um, there's actually a lot more that goes into, um, improving communication about data downtime, um, improving, okay, there's an issue, but what is the impact of that issue? Do I care about it? Who owns this? Who should start solving this? How do I know what the root cause is? And so, um, and how do I actually prevent this to begin with, right? So if we instill sort of the visibility here and empower people to see these things and to make changes with this context in mind, you can actually reduce these to begin with.

AI assessment note: “we actually use machine learning for that. Uh, so we sort of infer”

Q And out of curiosity, how does that translate, uh, for Monte Carlo into, um, selling? Like who's, who's your buyer? Who, who buys a platform?

A Our mission is to accelerate the world's adoption of data by reducing or helping to eliminate data downtime. And so that means that we work with data teams to help them reduce, uh, data downtime. Um, oftentimes the folks that we work with most closely are data engineers and data analysts. Um, Because they are mostly the folks who are responsible for data pipelines or for making sure that the data is actually accurate. Um, and working with their sort of consumers include data scientists or, or different teams like sort of marketing teams or, um, analytics teams are embedded within, um, of their business units, uh, who might consume the data. Um, so in this, in that case, for example, someone on the marketing team might have a question like, Which data set should I use or which report should I use? And, um, is it reliable? Uh, and so you might be able, you could use MoneyColor to answer that question. Um, but the primary sort of users for us are the data engineers and data analysts.

AI assessment note: “primary sort of users for us are the data engineers and data analysts”

Q All right. So the concept of data downtime leads to the concept of data observability. Uh, do you want to explain what that is?

A Organizations and, and data teams have invested a lot in their data infrastructure. Um, and we're seeing that in the rise of data infrastructure companies, right? So you're seeing companies like BigQuery with 1.5 billion dollars in revenue, Snowflake with a billion dollars in revenue, Um, Databricks with eight hundred million in accelerating. Um, and so, uh, organizations are investing a lot in building sort of best in class data. Um, and there are, you know, full teams, including data engineers, data analysts, data scientists that are responsible to actually deliver data products. And so, um, the responsibility of those teams is actually to deliver those data products in a reliable, trusted way. And that's actually really hard to do when the data is wrong often. Um, and so in order to kind of solve that, one approach is to actually look at, um, how is this solved in software engineering? Because software engineers actually have a similar role in making sure that infrastructure and web apps and other sort of, um, software products that they are building and designing are in fact reliable and are not down, so to speak. And so if you take that concept and you say, okay, how, what would that look like in the world of data? What if we took those concepts and apply them to data? And this is what we call sort of, you know, in quotation, the, um, good pipelines, bad data problems. So …

AI assessment note: “infer the health of a system based on its outputs”

Q Speaking of team members, um, how do you think about the sort of human and social aspect of data observability? Who owns this? Is that engineers? Is that business people? Um, and, uh, how do you think about it in the context of the emerging data mesh, which is, uh, something that I, I believe you spend a good amount of, uh, of time, um, thinking about.

A For folks not familiar with the data mesh at a very high level, it's sort of a concept that kind of, you know, taking the data industry by a storm, um, you know, love it or hate it. It's, it's very much used and kind of in discussion, but generally the idea of, um, making, uh, data decentralized and self-serve is something that we see a lot. Um, and that has to happen as part of data becoming widespread in the organization. So in the past, if you had only like two or three people Working with data, you can make it centralized, big deal. You could, you know, um, work with the data, check it, and you're good to go, ah, more or less. Today you have like hundreds of people working with the data. It does not make sense anymore that there's one team that sort of has the keys to it and, and, and, um, is really actually just ends up as a bottleneck. So, you know, my work with a customer was like, yeah, if I wanted to get something done with my data team, I basically have to wait a year in order for them to get through all of their priorities. Um, That's a reality for lots of data teams. Um, they have to wait months or years to get something done, which is just doesn't make sense for an organization that wants to really, you know, make data accessible for, for large number of teams do when it comes to ownership.

AI assessment note: “For folks not familiar with the data mesh at a very high level”