Episode 28: Unstructured Data Insight with RavenPack

Hey listeners, welcome to SphereCast, a podcast all about technology, technology advice, technology inspiration, and how real entrepreneurs have used technology to build their businesses from the ground up. If you're wondering how technology can support your business goals, rest assured our guests have been there and done that.

What if I were to tell you that there's a gold mine at your fingertips, a money tree? Too good to be true? Think again. The vast majority of the world's data is unformatted, not in spreadsheets, but in emails, not in columns and rows, but in news articles, tweets, and posts. And if you can find a way to sift through that data and act upon what you learn, the potential for success is great.

Hey, Luke here. And this week, I had the opportunity to speak with Kevin Crosbie, Chief Product Officer at RavenPack. RavenPack is a platform that gives meaning to unstructured data. A platform that offers speed, accuracy, in analyzing large amounts of unstructured content coming from newspapers, emails, social media, and more. The company then enables clients to quickly extract value and insight from the process data.

Regarding speed, RavenPack can analyze one page of data every 300 milliseconds, which is literally the time it takes to blink your eyes. Some estimate that between 80 and 90% of all of the world's data is unstructured, so the spoils go to the individuals and companies that find a way to interpret it. RavenPack has found a way to do just that.

On the podcast, Kevin discusses all the key use cases his platform can meet, explains the difficulties of migrating to the cloud, shares some of the most difficult decisions he's had to make, and describes what is it like to be driving the direction of an industry-leading product. So, without further ado, here's a conversation with Kevin Crosbie.

Hello, everyone. This is Luke from SphereCast and today, we are hosting Kevin Crosbie. Kevin, how are you?

Hi, Luke. How are you? Thank you very much for having me on your show. Doing pretty well. It's a nice day down here in sunny Marbella.

Nice. Nice. So, I'm going to start with that, you know, information, piece of information that you just shared with us. So, so RavenPack is based down in the sunny south of Spain. Tell me more about that. How did you end up there?

Yeah, Marbella is an amazing place to live. It's it's a tourist place with lots of mountains and beaches, great restaurants. As you might know, the weather is pretty good in Spain, but Marbella actually has a particular kind of a microclimate. So, that means that when it gets too hot everywhere else, it's it's still nice here or when it's too cold everywhere else, it's it's nice here. And I mean, RavenPack ultimately ended up in Marbella because our founder was just tired of large cities and pollution and traffic. And he had been to Marbella quite a few times. They've got good internet. They've got a good lifestyle. So, you know, everything that we do is based in the cloud these days, and that lifestyle aspect is definitely pretty much a winner for us when it comes to attracting good talent to come to work for us.

So, did you move to Marbella then or were you already based there?

I joined RavenPack back in 2003, but I was actually already in Spain. And I met the the founders through a mutual friend. And, you know, once I heard what they were up to, it was it was definitely something I wanted in on.

Okay. Okay, so you liked the project and then you're in completely. Okay, good. That's good to hear. That's a that's a good story. Okay, so I'm going to start with the most, you know, recent venture that we're talking about, RavenPack. So, in your own words, tell us, what is RavenPack and why did you join this company? I think you've given us some idea, but I'd I'd love to know a little bit more. What did you like about it and what appealed to you the most?

Yeah, so as I mentioned, you know, I heard what RavenPack was thinking of. The vision that was communicated to me at the beginning was you know, it's something kind of abstract, but it's like that there's there's so much information available in the world and a lot of it is in text format. And the way companies are addressing that is to have people read it, right? And it's it's very hard to keep on top of everything. And so many companies are looking or have been looking for ways to become more data-driven in the ways in which they they they find actionable insights.

So, that's that's how I how we started. That's that's what was interesting for me. For the company, when we did start, the obvious starting point was financial markets. And that's where, you know, information tends to be acted on very rapidly because investors are always looking for an edge. And I'd say that, you know, many of the the investment managers that that work with us would would operate in a fully systematic way and they're interested in not just looking at the information that they normally have available, which is things like stock price or volume of trades or fundamental information like earnings and revenues, but they want to be able to know when Elon Musk sends a tweet or when the president of the US says something that affects their investments in in some some way.

And and that's really how the the alternative data market was created, with what we call news analytics being one of the most sought after data sets, I mean, even today.

I see. Very interesting. So, okay, that I think that that's a good segue into the actual product. So, could you tell me more about the analytics piece that you are developing there? So, I've understood so far you're you're helping financial markets or the the operators in the financial markets, so tell me, can you tell us more about the product, maybe who it's helping and how how does it actually help them? So, news analytics is one part. Is Is Could you go deeper into this, maybe?

Yeah, I'll I'll go into that, but I'll I'll go into a little more detail about RavenPack and what it is we do. So, you know, RavenPack is a technology company and we're in the business of taking unstructured content, like text, news, social media, IMs, email, and turning that into strategic assets for companies to use by structuring that data. And we do that by using our own proprietary natural natural language processing technology. And what we try to do is interpret documents as a human might.

But obviously, you know, providing an output that computers can easily use and information that can be, let's say, onboarded directly into some kind of a process or even a reporting program, some kind of a business intelligence tool. And the way in which we do this, I mentioned like we try to look at documents as as, or text, as humans might, we track millions and millions of of named entities, and we track tag the stories that we, or the the documents that we receive, with the many layers of linguistic information that our system knows how to produce, like parts of speech, sentence structure, the actual mentions of entities or emotions, things like that.

We can also detect over 7,000 different market-moving events, you know, and also identify the role that each entity plays in that event. So, you know, just to give you some examples, for instance, companies can announce bankruptcy or they can appoint a new CEO, or they might one company might merge with another company. It's important to know who is who is merging with whom. Presidents can be inaugurated or presidents can be impeached. A a global pandemic can be declared or a pharmaceutical company might might produce a a vaccine that addresses that global pandemic. You could have a terrorist bombing for on an oil pipeline, and that could have, you know, effects on commodity markets.

So, there's all of these different things that can happen to the different entities that are mentioned in text, and you want to be able to stay on top of that and know what is going to affect you. So, you know, what RavenPack Analytics is, our our processes take text, and they do this analysis, and they ultimately enrich the information that we process with with all of this analysis, right?

So, there's a number of different product lines. The first one you mentioned is RavenPack Analytics. That is one of the outputs of our process. So, once we take the documents that we process and for RavenPack Analytics, that's kind of like our flagship product, it's news and social media from the year 2000 to present, and being updated in real time. So, there's a 20-plus year archive of information there. We provide real-time updates, and it takes us around 200 to 300 milliseconds to process one document. Just to put that in context, it takes 300 milliseconds for the human eye to blink.

And what we're doing is we're taking all of these documents that are coming out in the media, talking about the economy, talking about major events, talking about companies and what companies are saying. We're processing it using our our natural language language processing solution, and ultimately, what we do is we take that and convert it into a very simple data set that can be loaded into any relational database or loaded into any kind of a a a statistical analysis program, or a BI tool, something like that, right?

And not only do we detect things like the entities and the the events that are taking place, but we also give information on the sentiment of all of the mentions that we have. So, for instance, you know, you can have documents like, imagine a car company recalls some of their vehicles and they send out a press release that tells their customers that they're doing a good thing, you know, and they're going to recall this. And, you know, if you were to read that text, you would probably interpret it as something good because, well, it's it's it's gone through the hands of a good marketeer.

But, you know, our system knows that product recalls are terrible for companies. They usually cause stock prices to go down, investments to go down, and the company to perform worse because there's there's more, let's say, uncertainty about their their products. So, we supply all of these different views of sentiment that allow you to assess what is going on, and we we do it across a number of different different themes, as well as kind of a general score, there's an entity-specific score, the scores for things like sustainability, which is one of the main topics in in the world today with all of the the things that are happening to climate and climate change and new regulations coming out for companies.

We also provide information on things like relevance, how relevant the mentions are of these entities, and whether the thing that is being mentioned is something new or not, right? So, how's that used then? Ultimately, we sell RavenPack Analytics to, you know, anybody that is interested in in taking the media and making data-driven decisions, right? So, that could be the end users could be data scientists, or they could be business analysts, or risk officers in in corporations. Typically, our clients would be tier one banks and hedge funds, you know, on the investment side, you've got insurance companies, but it could be any corporation that's interested in competitive analysis, risk analysis, etc.

Normally, they're interested in looking at some of these entities and finding out what's going on over time. You know, they'll be interested in saying, "Okay, if I'm an investor, I've got a portfolio of companies, I might have 50 in there, and I want to be able to compute, based on what is being mentioned in the news, I might want to compute daily factors, let's say, that give me some kind of a trend analysis of what's going on from a sentiment point of view, or from the major events that are happening. I may wish to find out, you know, for the companies that are being mentioned that I'm interested in, who are the other companies that I would never have heard of before they're being mentioned in the news as well so that I can start to investigate them."

So, imagine if a company comes out with a new product. You might start looking at, okay, who are the competitors? Who are the new entrants in there that nobody's talking about yet? They're mentioned in the news because, you know, they probably come out with their own press releases and mentioned similar products. So, you could do that kind of analysis. So, our dataset is something that can be filtered very easily. We have lots and lots of different columns that allow you to filter over time, or by the entities you're interested in, or by sentiment, or specific events, ESG events, for instance, things like that.

And you can incorporate that into your into your into your models if you want to make these data-driven decisions. So, the way an investor might do that, for instance, you might come up with a model, let's say, that says, "I think that anytime a company appoints a new CEO, the next earnings report is going to be good because I think that the CEO is going to try to make the earnings good either by, you know, reducing costs, firing people, or by, you know, bringing on new partnerships and business that they had had on hand already." They want to make themselves look good, right? That's a good hypothesis.

You've a 20-year archive of every single company that has ever appointed a new CEO. You have your own, let's say, stock price information. So, you can start to build models that say, "Okay, what happens when the company next announces their earnings after a CEO is appointed? Is that a good signal?" And now you can build a model around that. You can assess whether it's good or bad, and then you can start it in real time, right? That's the kind of model that one of our clients might might might try to build.

But as I said, it's it's not just something that you you need to look at over long periods of time. You might just want to know what's happening today. If I look at my own company, who are my competitors and how are they being how are they being mentioned in the news? Is their sentiment better than ours?

So, that's that's the the kind of the the core RavenPack Analytics product that that we sell. We do have other products as well. I mean, we have that that's like pure raw data, and something that we sell as a data feed or as a as a large data set. But we also have interfaces that solve, let's say, productivity problems for, you know, risk officers again or investment analysts. They allow people to browse all of the content and do more smart searches to find and build connections between the things that they're interested in and what else is being mentioned in the media to discover, or what I always hear from our clients is, "Tell me what I don't know." That's how we do it. We tell people what they don't know yet by by bringing, surfacing these these relationships as they're being mentioned.

So, going to predictive there a little bit?

Yeah. I mean, the predictive part is is not what we're doing. We're just telling you what's happening in the in the the the news and social media right now. And and it's up to them really to to build their own models based on that or to analyze it, like it's as I said, this this would be more of a productivity tool for humans that can come in and read, build, let's say, watch lists of of of companies that they might need to start watching because they might start affecting their investments.

And then finally, we do have another product line where people can send us or not send us, but use our NLP service to analyze their own content and ultimately take things like their email or their instant messenger, which is something that they own as a corporation, they have been building, it's something that, you know, they've had since day one. Everybody's talking on this, but, you know, you have all these little silos of people that are not sharing information. And we allow them to upload this, have it analyzed in the exact same way, convert it into a structured dataset, which is ultimately a strategic asset now for them. And they can use that in order to make better decisions about, you know, what are the things that people are talking about right now, what is trending, what should they be watching, or, for instance, if they're interested in some new new concepts that a a research report that they're working on, you know, well, maybe if they search inside their own content, they can find who else is working on this and and start to collaborate better. So, there's lots of different applications.

Very interesting use cases there. So, okay. I have a bit of an understanding there. So, that's that's good. Thanks for sharing that with us. Okay, I'd I'd love to know a little bit about the product now. So, you have been a product officer in this organization now for some time. So, what role have you played in shaping the actual product and its roadmap? I mean, there must be a roadmap, I guess, and did you play a part in choosing technologies as well? Or was was it was there somebody that you collaborated with and in the end, you made the decision collectively?

Yeah, I've been in RavenPack since 2003, so I think that's 18 years now. Quite some time.

Yeah.

Yeah, too long. No, it's it's been fun. And I've I've I've really grown in the company. I started out as a senior developer, so I am very I have a technical background. And I took on the product role because I really liked the strategic aspect and meeting clients and finding out how people were using what it was that we were building. So, it was kind of natural that I stepped into that role. And yes, I would say that I have played a very big role in defining our roadmap, and also, you know, building all of the core components. Like it's it's a data set, so defining what the format of that data set is something that I would have played a very big role in, also the way we deliver it. We have a REST API. You can get things from file services. It's possible to use a tool that we have for building queries. And you know, there's there's been a number of different areas like that. I mean, I wouldn't be the the only person taking decisions like that. You know, we have a Chief Technology Officer who is, you know, an MIT grad. He's he's amazing. He he makes a lot of the decisions as well about where our technology would go.

Can I ask you something quickly? So, you have the REST API, but did you did you ever think about like because your customers could be using a number of different tools on their side, so how do you really integrate your solution with theirs? Are Are there any ready-made plug-and-play solutions that you provide them where you have built some, let's say, connectors or hooks that work on literally any tool they might be using on their end with the data sets?

Yeah, definitely. I mean, we have our REST API, which is kind of like the the the the easiest way to to work with our data, or let's say the the way in which anybody can interact with it. But we do have a number of different connectors, like Python APIs. It's possible to connect it directly to Excel or to MATLAB and R. They're kind of the core demands from our data scientists. It's possible to hook it up with Power BI and Tableau, those types of of tools as well.

ClickData, well, yeah. Yeah, okay. Great. Great. Okay, amazing. Anything else you want to mention about the, you know, selecting technologies? Obviously, you were there with the the CTO. So, I'm curious just to understand this. So, back in 2003, cloud hosting wasn't a big thing, right? So, and now it is. Now everything is on cloud. So, how has that evolved, if you can shed some light on that?

Yeah, definitely when we started out as a as a small company, you know, we used servers. You know, they were all actually on-prem initially. Then we moved into, you know, hosted solutions. Normally, we hosted right beside exchanges because, you know, there's a latency demand from some of our clients. They want to be able to get our data as, you know, before anybody else. So, we would put ourselves beside the exchanges where our clients also are.

So, yeah, I mean, that was fine when we were talking about, you know, maybe tens of thousands of documents per day. But now we're processing millions of documents per day and, you know, we have a historical archive that we often want to reclassify and just run the whole thing again. And yeah, definitely we became aware quite some time ago that we needed to move to the cloud, and so we did that quite some time ago. Everything that we do is now based in the cloud. We're heavy users of of many different components that are available. Definitely we're firm believers in in in build architecting as a service and, you know, serverless-type approach.

So, yeah, I mean, it's it's been quite a journey to to implement that. Definitely there were some growing pains at the beginning. Moving from a server-based solution to the cloud does involve quite a lot of re-architecture and we had to invest heavily in that. But once it was done, I mean, the benefits have paid off so much.

Amazing. Good to hear that. Are you Are you with GCP or AWS? Which one did you choose?

AWS.

AWS, okay. Right, good. Thanks for sharing that. I'm I'm going to segue, you know, this into something else. So, when you joined back in 2003, and now completely different things, you know. So, tell us about the main challenges. Let's Let's talk about the product officer role, right? So, could you talk about some of the challenges that you have tackled in this role or maybe something you want to highlight, some particular story that comes to mind?

Yeah, I guess, you know, one of the one of the key things is that when we started out and we realized that there was a demand for a product like this, there wasn't anything like this, right? So, we had to define it. We had to build it. And this market is something that, you know, we're definitely pioneers in. If you if you do look at the Wikipedia page for for news analytics, our CEO created it because there wasn't one, right? I mean, he created the stub for it and then people started filling it in over time.

But but definitely there's been a lot of challenges that we've need needed to solve there. So, for instance, you know, when we're when we're talking about tracking named entities, you might think of, "Yeah, you know, I I have an an entity. It's let's say, Microsoft or or Google or something like that." Well, actually, Google is now Alphabet. They went through some corporate changes. They were different over time. They changed. They acquired new companies, new subsidiaries. They changed their name. And we became aware that our clients, especially the ones that needed to do back-testing models over long periods of time, needed us to solve this what we call point-in-time problem. That when we're looking at a particular article that came out, let's say, in, you know, 2008, if we're reprocessing that again to to, you know, upgrade our technology, we have to look at that from the point of view of what entities were available at that point in time, who was really trading. You know, back then, it would have been Google. They changed their name to to Apple, though.

So, that was one of the the major problems that we had to overcome. There's also, you know, as as time has gone on, we've incorporated a lot of ML approaches. You know, you have a lot of companies that label themselves as AI firms, you know, just to kind of use the buzzword. We're we're not so into trying to label ourselves like that. I mean, when we see AI, we're talking about how to how to make it possible for machines to to to behave like humans, right? And ultimately, the the idea of of understanding text and understanding what's going on fits right in there, right?

So, initially when we came out with our product, we would have used a rules-based approach, you know, so something where, you know, we really required the context of what was being mentioned. Humans were building rules that were able to tag documents with events, let's say, you know, so you would have to see something like Microsoft announces bankruptcy in order to match that event, right? And as we went on over time, we also incorporated lots of ML approaches to learn new things, especially to do things like sentiment. But one of the things that we noticed is that, especially in the finance sphere, a lot of our clients prefer the rules-based approach because one of the issues you have with things like machine learning is that it's black box, right? You can't come and say, "This is why I got this result," or it's at least not very easy to do that.

And our clients require transparency. These are a lot of them are investment funds and they have charters that come out and say that they have to show the reasoning behind certain decisions. And if a solution comes and says, "Well, we can't tell you why," then that's not good enough. So, when it comes to, you know, providing information to the financial sphere, the rules-based approach is still a very valid one and one that we've put a lot of investment in. So, that's also been a challenge. Like, you know, do you turn over to ML completely? Do you not?

Also, from you know, from our point of view, a lot of people are looking like, give me one sentiment score. Sentiment is one of the the key drivers as well. We're we're very well known as a sentiment provider. And you know, our our approach has been, well, there is no one score. I gave you that example of the car company announcing a press release, you know, where they're being kind and benevolent by recalling some vehicles, but actually that's terrible news, right? So, we have a number of different sentiment indicators, and you know, knowing which ones are the ones to to kind of use in different situations is also something that, you know, has been a challenge to to get right over time.

I see. I mean, yeah, on my trading account with with every stock, I have the social media sentiment, sustainability score. I look at those things as well. I think they they do matter. But I did not understand until now like what exactly It's not for the common investor, though. It's more for the the the people that are sitting on the funds and investing in these companies on a larger scale, I guess. And they are the ones who are giving it a rating and that that's what affects the overall, you know, the day-to-day values of these things, I guess. So, hm.

I mean, some of those sentiment scores would be, you know, they're based on an analyst who is reading a lot of the ones that you get in investment platforms, like a sustainability score or a sentiment score, would normally be based on a team of analysts who are actually reading tons of content and coming up with They're looking at the balance sheets and the income statements of the companies and, you know, they're trying to get into the nitty-gritty. It's slow, right? I mean, they're probably a quarter off.

So, this is where our solution is able to give you things within milliseconds, right? So, so our clients can compute those kinds of scores in a very, very quick way, and incorporate information that's really trending right now.

That's amazing. I mean, I'm I'm a bit I'm a bit shocked, I should say. This is Once you start to dig into these kinds of things, you find out how things really work. This is on the edge stuff, right. Okay, that's great. Good to hear that. Could you tell me a little bit more about Let's Let's Let's talk about, you know, maybe one valuable, I know there have been many lessons, but maybe there's one that you still hold close to your heart, and every morning or every now and then, you remind yourself of some lesson that you have learned in the past in your professional career, or working at RavenPack?

I mean, I've been in RavenPack since the beginning. We started out as a small group of people working on on a problem, you know, and we built something that was definitely a product that we could grow and and, you know, general build a a successful company on. But today, the amount of things that we're trying to solve has just grown exponentially. And I think, for me, the most valuable lesson that I've learned across this is is to make sure that the team you have is amazing, otherwise you can't build products like this. We've We've got very We We definitely set the bar very high internally for ourselves. And and I think we're we're very well-known in the space as being the leader in what we do.

In RavenPack, you know, we've over 30 different nationalities. There's over 24 different spoken languages. We've got a strong culture. And you know, I'd I'd say that one of the things that really defines RavenPack is that the people that work there really feel a part of it, and they own what they do, and they're proud to to be doing it. And I think that's what really makes the difference. You know, having an amazing team, that's the that's the most valuable lesson.

And what what's the the idea behind the name, if I can ask you, if you know anything about it? Yeah.

Yeah, so RavenPack the name RavenPack comes from the the Norse god Odin. So, Odin had two ravens. There was Huginn and Muninn, and he used them He was blind or at least blind in one eye. And he used them to go out I think Huginn went out during the day and Muninn at nighttime, something like that. And they came back and they gave him the insights that he needed in order to be the most powerful god. So, we're the RavenPack. We're giving you the the the insights that are going to help you become the most powerful business in the world, right?

Oh, wow. I'm glad I asked you that question, so Most people don't know that. That's a that's a I was curious. I just thought this is a unique name, so there must be, and I I I quite often, I tell, you know, myself that a non-descriptive name is is often a good name because you have to It makes you curious just to learn more about the organization. So, yeah, that's that's really good to know. It's It's quite non-descriptive, but it has some meaning behind it, so that is very interesting. Yeah. Okay. Great. Okay, so I'm going to shift gears a little bit. So, let let me ask you something about decisions. So, as a Chief Product Officer, what do you feel are the most difficult decisions that you are making or have made or continue to make in in your day-to-day work?

Yeah, that's definitely that's my day-to-day right there is is is making difficult decisions. I'd say, you know, some of the hard things to do are, you know, prioritize. I mentioned that there's so many things that we want to do and we've got our fingers in lots of different pies because the world of of sentiment analysis, natural language processing is moving very fast, and we like to consider ourselves as being, you know, one of the leaders in that.

So, I'd say the the hardest part there is prioritizing. It's trying to figure out what is the thing that brings most value to our customers, identifying that and making sure that we don't just give a a basic solution. We We We give a a good solution that people are going to be able to use. And and I think that's why people look at us as being an industry leader here is because, you know, we've we've we pay a lot of attention to to the details.

I'd say, you know, some of the other decisions are on that I'd I'd have to do would be like, you know, typical things, like, do we build, do we buy? You know, building something obviously takes a lot of your resources and time. It's not free. You have to do it in-house. You know, maybe you buy certain things, like a database or something like that, and you rely on another company's solution. But that also has its own hidden costs. It takes a long time to onboard that.

Data is I've never seen a clean database after having bought it. It's always been something that we've had to put a ton of work into cleaning and making into something that we can use that has the levels of quality that we need. So, they they'd be kind of the I'd say the the key difficult decisions that that

Never-ending problems that are always creeping up. I see. I mean, there's an interesting episode on our podcast. I I spoke with Michael Cohen. He's, he's an ex-Uber guy, and he's working with an ex-Zoox guy, and they started a company called Spark AI. And at Spark AI, what they're doing is they are, let's say you have an algorithm which is giving you an efficiency of 90%. But for your customers or for you to be using that for the customers, that solution, you need 98%. So, what they are doing is they're taking in those exceptions from their customers into an API, so you can take your engineers, data engineers to solve another problem, and agree an SLA with them, and they will give you that result. So, you get your 98% and those exceptions are solved in real time very quickly. And you can have your staff, you know, doing the bigger problems or newer problems that you want to fix. And then you can revisit these at a later point or, you know, or learn something that, you know, that you were not wanting to invest the time in because it's it's too much of an edge case that you want to deal with right now, so

Exactly. And it's it's like that. It's like, you know, do you want to Is this your

It's the build versus buy, you know, thing.

Is it your Is this what you do? Do you want to actually be part of that or do you want to outsource that to somebody else?

Yeah. Yeah. Okay, interesting, interesting. Some threading there. Okay. Excellent. Now, I think, I was going to ask you this question, but I I will ask you, actually, because I was thinking of skipping it because COVID has become, you know, I don't think it's it's everyone has been impacted by COVID. There is no, you know, downgrading that sentiment. But I was thinking maybe we can skip, but knowing the industries that you operate in, so if you can quickly touch on this question. So, how has COVID-19 impacted your business or RavenPack? And what did you do to adjust to it? Last year, I think we saw a lot of chaos in in various organizations, people being laid off, etc. So, how was RavenPack doing then?

Yeah, I think initially, I mean, everybody was I don't think there's anybody that wasn't affected by it. Maybe Amazon was was the only one affected in a good way, right? I mean, because their their revenues definitely went up.

Initially, you know, I'd say probably the the the key thing was just, you know, companies implementing risk mitigation plans where they just put a a hold on buying, right? So, you know, there was a certain amount of time where it was just difficult to get people to make decisions, right? That would have been last year sometime, but, you know, I think, for the most part, the the pandemic hasn't affected RavenPack itself. From the point of view of the workers, you know, and how we work, we were already set up to to work remote work. We were able to do that overnight.

And from a business perspective, you know, even though we had that initial kind of moment where people people were were were putting a hold on things, I think the pandemic actually made firms realize that that they actually need to make more data-driven decisions and that they need to be able to monitor the media for important things like this that are happening. Like, you know, if if you're a a a customer of a bunch of different suppliers, you probably want to know if those suppliers are going to be affected by things like pandemics and how are they being mentioned in the news. What kind of trends are happening, you know, in certain areas, like is there is there a big drop in in purchases for for your product segment, you know, that kind of thing.

So, if anything, you know, I think, over time, we've had more interest in in, in our type of product. Definitely, people have looked outside of kind of the standard use cases, and they're applying it to new areas. And one of the things that we actually did, I'm not sure if you're aware of this, but we created a a public freely available public website. It's called the Coronavirus Media Monitor, and it's something that we were able to produce to try and help people trawl through all of the the media that was going on. You know, there was a lot of panic. There was a lot of media hype that was taking place. People were worried. They didn't know what was fake news, what was not, what to listen to. We had constant repetition of the same things. Some of it was was false information, you know.

So, what we did was we put together this this media monitor, and it allows you to track all of the the major things that are happening on a global basis. You can see it like, you know, on a geographical map. What are the impacts of COVID? You can drill down into Spain, you can see what the infection levels are and compare that to what the perception of media panic is in that that country, and also, you know, what are the kind of key trending topics? Are people worried about things like lockdowns or, you know, not being able to purchase something the next day because there's there's there's some kind of a panic buying going on.

So, yeah, I think COVID impacted the world. I wouldn't say that our business was heavily affected by it.

I've just jumped onto the website and I'm definitely going to play with it a little bit to understand more about coronavirus media.

Yeah, we've had tons of people ask us for that data actually, and we've we've given it out. And we've had a lot of a lot of different research reports built on it. How could you use data like this in order to make have made better decisions? So, it's it's actually sparked a ton of interest, but, I mean, we produce that in order to to kind of do something that could help. I mean, that was that was the main reason for it.

Yeah, very social of RavenPack to do this. Thank you. I'm I'm I'm going to check it out after that and send you an email. Right, okay. Right, let's switch gears a little bit. So, I'm going to go towards the personal side. Kevin, can you describe your typical day, if there is a such thing nowadays? I wouldn't think so. I mean, I my day is completely different. Every day is is new, different challenges. What about you?

Yeah, I'd say you're right. My typical day is not very typical. I do lots of meetings, I'm sure, like most people in my position would. I'm I'm managing quite a large team, but I have some amazing direct reports who are able to to to really stand up and and help me out. I think my main day-to-day would be to to just make sure people aren't stuck, you know, to help people make hard decisions about what should happen in our product.

Like, I mean, talking things like really fine-grained things, like maybe solving a problem with how to detect a unit that's mentioned. Imagine if a story comes out and says WhatsApp CEO resigned in 2000. Well, do you attribute that to WhatsApp under under Facebook or not? You know, so we've got all of these decisions that we need to make. And, you know, what I try to do is just make sure people aren't stuck.

The team is big, but we're constantly hiring, so you know, I spend a long time hiring, hiring, hiring, doing interviews, making sure my team are also able to to pick things up there. And ultimately, I mean, my main goal is to to make sure that I'm making an environment where where the people in my team can be successful, that they can they can build the things that we need to build in order to be an amazing organization.

Got it. Do you hire in-house only or are you working with some, you know, hiring companies as well?

We we do whatever we need to to get the right people. I mean, yeah, we we'll freelance if we need to for for some tasks, you know, we've we've definitely outsourced quite a lot of a lot of work. We're firm believers in that. We in-house for a a lot of our staff positions. We do engage with hiring companies when needed as well.

Awesome. Okay. Right. So, this is a tricky one, so but I have to ask you. This is part of the process. So, tell us about one thing you wish you had known at the beginning of your career or when you started off back in 2003, since I'm understanding that, you know, you've you've spent most of your career here at RavenPack, so any anything you wish you had known?

Yeah, I guess, you know, if I were to to try and answer that, it's it's a tough question, all right. You know, I did start out as a as a developer, so, you know, that's one aspect of my career, but I think the way I look at myself today is as a as a a product person. And, you know, when I started out, I was the only person really running product. It was me and the CEO and the CTO really, you know, pushing things along. I was doing things on the product side. And it took a while before I realized that I needed to have smart people working with me. And that's why I mentioned my my direct reports. I have some amazing direct reports. They're super smart. They're able to work completely autonomously, solve problems whenever needed. So, I think that's that's the the best thing that I I I should have known sooner.

Yeah, they say that if you're the smartest one in the group, then there's something wrong with your surroundings. Right. Okay, yeah. Kevin, how how do you, I know we're all busy, but I think you are super busy. You are managing a large team. So, how do you manage how are you managing your time and how are you getting so much done? Any secrets?

Yeah, I I think, you know, the the main thing is trusting my team to get things done. That's the that's the way I manage my time. You know, I try and help them, try to make sure they're not getting stuck. You know, I don't try to empower people. I think people should be autonomous and and just, you know, operate themselves and take responsibility and be owners. So, you know, I do a lot by just trusting my team to to get things done.

I I'm a firm believer in using calendars, so my days and my my weeks are usually pretty well plotted out in advance. But I think if I were to say like, you know, how do we get things done, one of the one of the things that always, you know, it bites people, I think, is is when you leave things too late, you procrastinate and you try and solve things at the very end. It's it's student syndrome, I think they call it. So, I think one of the big things is avoid student syndrome, try and get things done as soon as possible, or you know, there's always going to be things you have to leave until the end, but I think the the idea is start on the hardest problems first, you know, try and get those unknowns out of the way as soon as possible.

Yeah, I completely agree. In violent agreement, as they say, with that one. Student syndrome, I suffered from it largely throughout my uni days, you know, leaving these assignments to the last few days and then there are some sleepless nights, you know, and, and then you suffer for a few days. And yeah, that's it. That's the story of a student, you know. Yeah. Okay, never mind. We'll move on from there. So, thanks for sharing those tips. How do you unwind, Kevin? Any any particular things you do? I know you have a you have two daughters, so do you get any time to unwind, actually, or is

Yeah, I was about to say that, I have two children.

Two daughters is unwinding for you.

I come to work to unwind.

I see. Yeah.

I I I mean, definitely my family is is the best way to unwind, spending time with them. When you are playing with a 3-year-old and a 5-year-old, I mean, you have to put work out of your head. You have to disconnect. You have to to move over and take them in because they don't take no for an answer. So, I think that's the that's the best way to unwind.

You know, there's there's we interviewed another CTO a few weeks ago. He He's CTO at a company called Dizbo. So, we asked him, "How do you unwind?" And he said the same thing. He said like, "Work is my unwind." And I said, "How? Explain it." So, his his story was that when he was like very really young, he read in a Japanese book that you make work the easier easiest part of your life by having other challenges around it. So, so this CTO is basically, when he's, so his name is Kwaku, so when he's not working, he's mountain climbing and having other massive challenges. So, when he comes to work, it's like, you know, walk in the park for him. And that that's his way of doing it, but, yeah.

That's a wise man.

Yeah. Okay, so we've come to the final part of our Q&A today. So, I'd love to know, you guys are using some leading, you know, bleeding-edge technologies and probably, you know, RavenPack is one of the trendsetters in this area. So, I'd love to know, how do you stay on top of latest technology innovations, or how do you innovate in an industry where there's where you are leaders? How do you keep doing it so you're not so you're not left behind, or keep ahead of the competition?

Yeah, I mean, we are, as you mentioned, it's we we are innovating ourselves. We're we're driving what the the leading edge is. So, you know, from that point of view, we're we're staying on top of it by defining it. But, you know, I I wouldn't say that's that's everything like. I mean, definitely we're we're always looking out for better solutions, you know, new technologies. We have got very direct relationships with our providers, like Amazon, for instance, and we get some great help from them, you know, whenever we need it on, you know, what what what are the latest and greatest that they're working on, how can we use them more.

You know, with with other things, definitely, knowledge sharing is is something that we try to promote across across the whole team, making sure that, you know, we we use our own instant messaging, chat system internally. And, you know, anytime somebody discovers something, they make it available, and you know, everybody gets to to to read up on it. We try to share that knowledge. And, and I mean, yeah, you're right, we process the news, so I we do use our own tools for this quite a bit as well.

Oh, great. Okay. Interesting. Very interesting. So, just, you know, continuation of the question I asked you, so what what continues to be the the biggest challenge from a technology perspective for you? And how how do you think it could be solved better or made more efficient?

Yeah, I think I guess the the the challenge that everybody has is that there's more and more data. It's growing exponentially. And as data gets bigger, not to use the the buzzword big data, but as as there's more and more data out there, you know, it does become more challenging to to be able to to operate on that. Like I mentioned that we we reprocess our whole archive periodically, you know, we need to be able to do that in a in a a timely and cost-effective way.

So, you know, there's there's definitely big challenges out there in processing data. What we try to do for our clients is to take that heavy lifting away from them and to to do all of that hard work and give them something that's much simpler to work with. And that's that's how we're trying to to to be efficient there.

Yeah, I guess that's that's USP here, isn't it? The you give them the insight, that's RavenPack. That's mythology behind, so. Okay. Right. Okay. So, with that, we have come to the last part, and it's about leadership. So, I'd love to know from you, what do you think are the top three attributes, I I think there are many attributes, but, you know, if we were to choose just three, top three attributes of a great leader in current times?

I'd say, you know, by current times, if we were to define that, unfortunately, we have to define that as being the pandemic, right? Which is post-pandemic.

Post-pandemic or in in the middle of it, I agree.

Yeah, and I think, you know, the last year and a half have really taught taught us a lot of lessons, you know, on how to act under this type of situation, you know. And I think some of the the key things that a good leader needs to to be able to do is to make sure that you keep people motivated. You know, give them a purpose, share the vision with them, make sure that they're on board with that and that everybody's working together to to get there, to get to the same place.

Also, I think, you know, there's because of what's going on, a lot of people go through a lot of stress and worry and there's a lot of feeling of despair sometimes. And I think that can be addressed or, let's say, lessened by just making sure people are informed, that they know what's going on, giving them some insight so that, you know, they're not feeling like they're in the dark. So, making sure you're a strong communicator, I think, is important.

And then, you know, if if you are a business that's trying to act in in any times, not just the pandemic, I mean, the pandemic, I think, what it's done is it's it's taken us to an extreme. When you take things to an extreme, you get to see certain characteristics and behaviors that you wouldn't normally see in the status quo, right? So, that's all we're really seeing with the pandemic. And if if you're a business that has a strong core with a reliable model, then you have a much better chance of surviving and prospering if you're if you have that behind you, you know. And during the pandemic is something that will really push you and and and kind of verify that or validate it. So, you know, I think business integrity is is another another strong point that a great leader needs to have.

Yeah, very wise words there. So, yeah, thanks for sharing that. I think the communication part is really important. I was thinking about it a few weeks ago. Let me tell you something, give you an example. So, in my like when I was a student, let's say, you know, 13 years ago, I finished my uni and while I was in the uni, I I started working in a in a place, an energy company based out of the UK in the North East. And what I found there was like, if you did something wrong, or if you made a mistake, there was this period that they would just keep going on and on and and I've Looking back, you know, at my career, I would think I would never do that again because they will just keep you go going through that despair and worrying period, which I didn't have to go through and there was so much stress. And imagine how that changes my mental makeup going into new jobs or other workplaces that I've been at. So, I was thinking about it. So, communication is really important, how you treat your colleagues. And I think this is something that a lot of the old corporate organizations don't realize how their policies affect the mental side of employees and when they're when they're communicating badly. So, and I I I really think, you know, for that reason, I would never go back to corporate. I think I'll I'll stick with startups or scale-ups for that reason. So.

Yeah, I think I think that's that's also wise words. I mean, it's we've definitely tried to to maintain our startup mentality as we've as we've grown.

Okay. Great. So, Kevin, with that, we have come to the end of the episode. I want to thank you for speaking with us. I I really enjoyed it. We've We've got good part of an hour here done and we were expecting to finish in 35 minutes, but it took like 55. So, thanks for thanks for being patient, and and I think it was really enjoyable speaking with you. And we hope to host you again in in 6 to 9 months period. So, I'll be sending you an invite again. Hopefully, you can join us.

Yeah, it was definitely a pleasure to have participated, and it was definitely a very interesting conversation. Thank you so much, Luke.

A special thanks to our podcast guest this week, and once again, to our sponsor, Sphere Partners, for bringing this episode to life. If you enjoyed this episode, drop SphereCast a five-star review on iTunes, and share this content with your network. For any relevant links or notes from this episode, check out our podcast website at www.sphereinc.com/spherecast. And always remember, when you think you can't, technology can. See you next time.