Welcome to our newest season of HumAIn podcast in 2021. HumAIn is your first look at the startups and industry titans that are leading and disrupting ML and AI data science, developer tools, and technical education. I am your host, David Yakobovitch, and this is HumAIn. If you like this episode, remember to subscribe and leave a review.  Now onto our show.

David Yakobovitch

Welcome back, listeners, to the HumAIn podcast. Today, I bring to you Geoffrey Horrell¹ who’s the head of Refinitiv Labs² from London, within the London Stock Exchange Group (LSEG). Geoffrey and I have connected through Refinitiv as a global player in data, APIs and the alternative data space. Many of our listeners know that I’m very big in both the FinTech and the data space. I’ve actually attended with Refinitiv their developer days in New York city, as well as the open source strategy forum, where they participated with open source data science technology.

Geoffrey, thanks so much for joining us on the show.

Geoffrey Horrell 

Thanks, David. Great to be here. 

David Yakobovitch

Well, I’m really excited to hear about where Refinitiv It is today. There’s a lot of exciting changes in 2021. Can you tee up for our audience where Refinitiv is today, Refinitiv Labs and your growth with the London Stock Exchange Group?

Geoffrey Horrell 

Everything is a very interesting time. What you’re seeing now is that we have Refinitiv, which was a global provider of data and workflow solutions. And, as you said, APIs and something, hopefully to reach out to the developer community who wanted to get more data into their applications and drive their strategies within wealth management, investment management, trading, risk, all these different sectors that we were serving.

Then, I guess, there was an opportunity to come together with the London Stock Exchange Group. Where they add an extra dimension to that data and workflow business. Where, obviously, people are familiar with the London Stock Exchange, running a venue. So, companies listing on the exchange, trading on that exchange.

There’s a great demand for people to see, well: how can I access? Not just the equity market, like when we talk to stock exchange groups. But also, the fixed income trading market, which Refinitiv provides, and FX treating marketplace, which Refinitiv provided. So, though you can see these different elements coming together, and then that’s on the kind of research your trade with the data, trade on the market with those solutions.

Now the other part of the stock exchange group is the London clearing house, which is the kind of post trade. So once the trades have been done, making sure things settle up and are cleared. So really we’re now serving the all financial ecosystem in one company, which it’s exciting from a market service and a growth opportunity. But as somebody who loves data, I’m a data geek at heart. All that amazing data that we now have access to within the labs in particular, we’re very excited about. So being able to build new kinds of analytics and really serve as customers at each part of their investment life cycle, is something that’s keeping us in the labs very busy and  keeping us really excited.

David Yakobovitch

I find that very exciting because knowing the history of Refinitiv, being a former Thomson Reuters company and being focused on both North America, Europe and the global markets. It’s almost coming back to the roots and growing from there, seeing how many people in the financial world know this; but not everyone else does is that London is still the center of finance.

When we think about Brexit in the industry, everyone says: London’s done. I say: No, it’s not. I was just in London before the pandemic at Hyde Park in January, 2020. My last major trip before the pandemic. Business was booming and it sounds like now is a great time for continued global expansion.

Geoffrey Horrell 

That’s right. Even though it’s called London Stock Exchange Group, we’re serving customers all around the world. In fact, London’s a great place to be, because you kind of have one foot in the Asia time zone and one foot in the North America time zone. I spend a lot of time with our colleagues in Singapore. We have a lab in Singapore and those fantastic growth markets there. And one of the things we can do now is bring some of the assets that let stock exchange group has: FTSE Russell indexes, for example. Bring them together with the Refinitiv data and that sort of reach into Asia. Actually our footprint is kind of equal actually between Asia, Europe and the Americas. But London is a great place to be and I am not going anywhere, and we’re still here.

David Yakobovitch

Thinking about technology. It’s so fascinating, as you described, that these labs are everywhere and it’s an exciting time to be part of labs. I recently got back from a multi-month trip at the start of 2021, where I was in Taiwan seeing how Asia has grown with data and technology and privacy as a result of the pandemic. Prior to that, had been involved with different evidence-based labs and data science labs in the United States, seeing how the world is one ecosystem, whether remote distributed or in person. So taking this back to Refinitiv Labs. What’s cooking today, what’s happening with new projects and technology? 

Geoffrey Horrell 

There’s labs, different labs, doing different kinds of things. So some labs out there are really partnering with FinTech to incubate them and grow them, and things like that. Some accelerator type functions. We don’t really do that as much in Renfinitiv Labs. We do have fintech partnerships, but our focus in the labs is partnering with our existing business units and bringing to them our experience and knowledge of. I just said what’s happening in the global world of technology.

So, big tech. We talk about all the cloud providers, all the different tools, machine learning tools, new ways of managing data. And as incredibly we see something like the tool set from LinkedIn, like KAFKA, and that’s designed to help manage streaming data for that use case, saying: Hey, why don’t we bring that in and can that help us with some of the work we’re doing with streaming data, with streaming financial data? 

So we sort of bring some of that open source world into sort of core bits of our business, so that education. Also, one of the things we partner with our business units on, just thinking differently about how you break down a problem. That’s the kind of full service part that you have user experience. We do customer research and we bring that lean startup, which probably many of your listeners are familiar with this approach. Which is to build something rapidly. Just test it with a few customers and iterate wrap, quickly.

The financial markets isn’t the kind of place where normally people do that, It’s highly regulated. People are very cautious about taking new things out, but actually what you’ve seen also is our customers have changed. So the customers expect great customer experience in their applications, even if not that not consumer applications or kind of enterprise applications, they still want to have a great user experience fit with their workflow.

Of course, increasingly have really rich, personalized data experiences within those applications. So as an example, some of the things that we’ve done in the labs recently, we focused on within the buy side; big asset managers overwhelmed with research documents and got to read all this research process to understand what’s going on. 

So the question is: Well, can we make that work flow better? Can we break down for themes that are driving the underlying research and can we score them with sentiment so that you can kind of see what’s important? What’s not important? And so we’ve built within the labs, a completely new workflow around that called SentiMine. And we’ve been taking that out and testing that with customers at an early stage. And then, based on that user feedback, we sort of pushed that through into our normal product launch process. That was done out of our Singapore lab, great project. In our London lab we did some similar work where We’re trying to deliver data. But this growth in the data science community and that maturing of the data science community, like: How do we reach this important customer base in a different way? And so, again, we’ve said: Well, how did data science work? What do they need, what they care about? 

So we built a capability called The Data Science Accelerator. Which mixed large sample data, large sample sets of data available with tutorials, with examples, with Jupyter notebooks with kind of on, easy… we don’t have to worry about compute or a cloud access or anything, It just works. So one of the barriers we found within the scientists was they want access, not just a small sample of data, but they want a lot of data that haven’t come from the financial community. So they don’t know how to use some of that financial data. So, again, bringing that together, so that they can kind of test their hypothesis short to their boss and go:  Hey, I really need this dataset. And I can prove to you why it works before you’ve had to go through a whole lengthy trial and provisioning and an environment provisioning process. 

So, both, thinking about workflows and also thinking about changes in our business model of how we can have reached customers. So they can kind of benefit from this sort of high quality of financial data. So all kinds of exciting things are going on in the labs right now. 

David Yakobovitch

I love how when you’re framing this, Geoffrey, on workflows. You can think about the start of a project: planning, experimentation, organizing data for augmentation with the data science accelerator. But then taking that further with pipelines using technology like Kafka and Sentiment Analysis. With your SentiMine solutions you’re bringing everything together, whether they’re starting out as experimentations or projects that are in development; they can become full products for the marketplace. 

You’ve seen on first hand so many of these products, both in-house with proprietary technology, as well as open source technology. And one thing I’ve seen as well, in 2020 and 2021, that data science as a service is going mainstream. The tools and technology are always changing, but the strategy is still consistent for workflows and we’ve seen data scientists respond and share what they’ve seen as some of the latest and greatest.

We were talking before the episode about one of the surveys that you ran last year to discover whether data scientists think is important for them. Can you share with our audience a little bit about why Refinitiv ran this survey with data scientists, your customers, your partners, and let’s start unpacking some things that were discovered?

Geoffrey Horrell 

It’s such an interesting thing to do because we’re putting the survey out as practitioners. So part of it is like: we’re just curious about what are sort of colleagues in the data science industry in finance doing. But it’s also so fascinating to see how in consumer retail or in other areas or ad tech, data scientists have been around and doing great stuff for a long time, but in financial services it’s a different kind of story.

That evolution of big firms would have quants and quants teams for risk or for, particularly, functions. But the last two services we’ve run, were just in a huge growth in the number of data scientists coming into the industry. We ran the survey both to understand like: What’s happening about the role itself, but also what do they need?

As you said, it’s an emerging industry. As an emerging capability. So, what do people want? What services do they need? What tools do they need? What kind of data do they need? What kind of projects are they working on?  So, what we’ve seen in the headline of this is that data scientists within the financial sector are really on the rise in a big way.

The numbers have grown 260% increase since 2018. It’s huge. Also a shift in terms of: where are they? So they’re moving from centralized, like shared functions, to be embedded deeply within individual business units, so really partnering with the businesses. Different business units within finance. So we’ve seen a few shifts within that growth.

David Yakobovitch

When we think of this workflow. Your new research report is coined the rise of the data scientist. I recently ran some workshops on serving ML and AI, these modern data science workflows and we have core titles that are now participating in this end to end workflow. You have the data scientists, the data engineer, the machine learning engineer, the software engineer, the product manager, even the site reliability engineer, but all these threads go back to the data scientists. So can you tell us about the maturing role of a data scientist in the firm and why now are we seeing the rise of the data scientist?

Geoffrey Horrell 

So you’re seeing the rise, are you saying in a couple of ways. One is that there are more and more business units. So when I say that, I guess I let me break that down. So the different use cases. So, as I said before, market risk, credit risk, those are areas that traditionally had quants and sort of senior analytics managers in those things.

What you’ve seen is other functions, reporting and compliance, portfolio management, investment research, idea generation, trade execution,  pre-trade. So we serve, there’s like a dozen different types of use cases where they’ve said. They’ve realized both the technology is mature. The techniques are mature. But also the industry has said: We need to get an edge in these areas, either for efficiency or to generate alpha.

So this brings data science into that. What you’ve seen data scientists having to do is not just crunch the numbers and build models, but also advise how you should set this project up? How do we break down the business problem on the one side? So that kind of strategic direction of like: How do we do this well?

The second part is, how do we set up ourselves for success in terms of, we’re not just going to do like one model and be done with machine learning. You’re going to have to keep iterating over and over. And you’re gonna have to do monitoring and reporting. There’s a governance element. There’s a technology pipeline element to it.

So per data scientist is being stretched in all these different directions to provide this advice. As you said, what you’re seeing then is that the strategic role of the data scientist is not just in how do I build this model, but it’s also in how does a company set themselves up to understand the end to end flow around it? So that’s the rise, in terms of numbers, but It’s also rise in terms of the strategic input that data science experts are having on the direction of firms.

David Yakobovitch

The classic dilemma we see with data scientists is where do they fit in an organization being involved with different startups and scale-ups? You’ll see some teams that have one data scientist and some teams that have a hundred data scientists. There’s often these different modalities of running and scaling these teams from being centralized to hub and spoke to ad hoc and different plans of attack at every organization. I don’t think any model is the best, but they’re all really fascinating in how different teams in different industries deliver products at scale. Have you seen any trends on that for teams and organizational functions from this year to report?

Geoffrey Horrell 

We have, in fact, that one of the stark things  that lead jumped out when we saw the numbers, is that the distribution. So that centralized model where it’s like: Hey, let’s have a big centralized analytics team. All the data scientists here, that’s shifted. 

So in 2018 report or 19 report, average number of teams. So we surveyed 420 different people across the industry. Everyone, the buy side, sell side over a billion dollars in revenue. So across all those kinds of firms, which I guess are slightly larger firms. The number of teams,  the number of data science teams grew from 2.7 to 7.1. So much bigger distribution. And that’s driven, because the data science capability it needs to fit with the domain knowledge and expertise within those different functions. So I was saying, what you need for pre-trade execution is very different for what you need for risk and compliance.

So data scientists being embedded within those groups. That’s the model that we’ve seen. But, just like you said, with data and data governance, where you would have maybe a distributed people working on data, but you’d have like a chief data officer with the kind of governance and oversight view.

You have the same model emerging where you have kind of distributed data scientists in different teams, but you maybe have a centralized, like analytics governance function, model governance function, something like that. So there’s some centralization remaining, which could be tooling or compliance on that kind of stuff. And then you’ve got the kind of embedding in the different functions that kind of distributing.

David Yakobovitch

So thinking about all these functions, I spit balled some titles before: data scientists, data engineer, machine learning engineer, software engineer, product managers, site reliability engineer. Are you seeing any growth or evolution of some of these newer or more traditional roles that are supporting the infrastructure and data scientists? 

Geoffrey Horrell 

Absolutely. Definitely on the engineering side that the ML operations. It’s funny because operations is kind of a dirty word in engineering, even though that’s the stuff that is required to make sure things work. But the sort of ML ops or ML engineering, definitely we see a growth there. 

Specialization there it’s difficult because you’re trying to get somebody who understands enough about data science models and stats and governance, but also is spending all day everyday or the engineering side. So that’s a really interesting hybrid. It’s massive. There’s a big shortage, actually, in the industry in financial services in data engineering.

It’s a really tight market right now for data engineering. Because it’s that bridge between understanding the data and the domain. Understanding data science and understanding the database selection or the normalizations that need to be applied. So data engineering, absolutely, huge growth. We think we were seeing there in the survey responses.

Also it is another shift, which is the analytics management. So the sort of senior management who are kind of setting up these groups. They’re stretching their knowledge and their skillset, more technical folks moving into those kinds of roles. So definitely some development and, also, like traditional roles like quants. 

So, in financial services you’ve always had quant research.There’s this dilemma around titles. So if I’m a quant, do I stay as a quant or a financial engineer? Do I rename myself as a data scientist; is that the right way to go? Or actually as the data science market matures, people go back and go: Actually, you know what? That is a quant rule. That’s why we need quants for this.

That is a visualization and business analysis role. So let’s not call that data science, let’s call that what it is; then within data science, the specialization: I’m an NLP expert,  I’m a vision expert or I’m a time series expert. Those specializations that you’re going to see within data science. 

It’s a really fluid time for some of these roles. It can be confusing in the market, like for people trying to build their career. But what we see is certainly the data scientist developing and maturing the strategic route, the engineering route, the data route, that sort of business engagement. So there’s a lot happening, David. So I’m just not going to keep talking. But you and I, both, are fascinated by what we’re seeing in this market. 

David Yakobovitch

As you mentioned Geoffrey, the data engineer is an essential growing role that’s in partnership with the data scientist. At at Single Store we’re building the next generation platform for scaled analytics, pipelines streaming for all analytics and transaction workloads. We use technology like Kafka, that you mentioned earlier in the show. 

We help minimize the infrastructure for DBH, so you can quickly spin up using our streamlined Terraform, Ansible, K8S (Kubernetes) and any of that tech to very quickly get from code to insight. 

I’ve not only seen that at our work today in SingleStore where we help that the Super Bowl run in real time or help the ads that you’re served in your media platforms or gaming platforms make personalized experiences. But, like you said, it’s across the whole industry. With the analytics we’re seeing with quants, just such a rise of new data and alternative data sets on both the buy side and sell side.

The challenge is, well, we think about ethics and ethical AI and human AI and human centered AI. Where does this lead us to about the best approach for building strategies and building products? Do you think we are moving to a place where we’re going to have more standards or a more experimental nature?

Geoffrey Horrell 

There’s both regulation and the threat of regulation that is going to come through around in these areas. So of course, GDPR, which is huge in Europe and has been a massive shift around how do you manage any data set that has any personal identifiable information. That’s critical, but beyond that, the ethics of AI where, It’s not just issued my model fair; but the role that did it engineer, actually, to come back to that, is also: Do I really understand my data source? Where has it come from? Has it been sourced in an ethical way? So we’ve seen people scraping data. It’s not licensed. You’re scraping data off of websites and it pretty clearly says: you do not script this data.

Do you really want to build your strategy on that data? Or perhaps I’m pulling data from an alternative data provider who is , actually, monitoring individuals and their traveling and worthy go. And again, has that been anonymized? Is that an ethical thing to be doing? Right through to when we do a lot of work with obviously Thomson Reuters, our texts.

So our news or filings or research documents; huge, massive, professionally curated sets of unstructured data. But then if I apply a standard sentiment tool on that, how biased is that a sentiment? Does it account for gender? Does it account for language differences? Actually, if you look there’s a lot of interesting research on how sentiment, and so on, is hugely subjective. That’s an area where there’s a lot of future interest, but the regulation is going to come on. Things like lending and consumer retail and credit, and it already is, but will continue into all these other areas.

So actually we put out a paper around alternative data and considerations around alternative data sources. One of the things we always say is we’ll come to a definitive and gap. If you’re like fully approved, signed off clean, clear, data sets rather than scraping a bit here in a bit there, you could come to a vendor like us and get actual professional data sources. That there’s a role there for the data engineer in terms of sourcing and decision-making about what data they want to bring in. But then also there’s a growing need to respond to not just governance. It’s here now, but anticipate. The kind of governance is going to be coming.

David Yakobovitch

To even anticipate, whether you want to buy a license or build data and data products. Of course, this is the classic age old problem that we see about in software technology. Do you build or buy, and that’s where solutions that Refinitiv provides helps you minimize the need to build some of these solutions. You can buy them off the shelf to scale, but it also means that traditionally the purchasing power at the buy-side would often be at different managing directors or in different tech companies with the CIO or certain heads of divisions. But now there’s this movement of this power and autonomy for data to the data scientists to say: You can be a decision maker, you can have added pressure to drive our strategy to be data driven.

Geoffrey Horrell 

That was one of the things that really surprised us in the report, because one of the questions we asked was. The traditional market data manager, as they’re called, is a purchasing or vendor sourcing person at a financial firm, they would normally be the one signing the contracts and doing the licensing and so on.

And, what you’ve seen, because data is the differentiator and effective management of differentiation. It is differentiating either in reducing costs through automation or in terms of driving trading strategy. What you’ve seen is the data scientist being the one who’s taking the lead in evaluating, in testing data. Scientists are saying 83% of the time they are the ones who are involved in trialing the data. But over 50% of the time, they’re also the one who makes the final decision in the data. At the time they were involved in a third of the time, they are the one who makes the final choice.

So that’s a big shift from where it was a couple of years ago. So very much, as I say, that rise in strategic importance of data scientists in terms of making these final calls on the data. 

David Yakobovitch

There was a big report that came out a few years ago from McKinsey. I remember it said along the lines that by 2030, 70% of companies will be AI first companies in some capacity using a part of the AI technology.

I also remember that when this report came out, this was around 2015, 2016, only about under 10 or 15% of companies said that they’re even exploring AI technologies. Something very shocking and supporting research from your report on the rise of data scientists shows that these numbers are getting closer today, even though we’re not at 2030.

Geoffrey Horrell 

The acceleration has really taken off in the last  24 months of the last few years, where everybody’s been at home and thought: Right, hang on, I can’t go visit my customer. I can’t make the same decisions that I would have made in-person with my team. How do I get the data together to make those choices?

And that’s even accelerating that full end to end digitization. So if we do this when we do the survey next year, we’ll see it even move even further. But the survey said that 72% of the businesses we talked to said that ML is a core component of their business strategy.

And when you see it, it’s not just the paying lip service to that, because they’re actually following up with investment in the technology, tools, investment in the people, and in the distribution, as I said, into the different business units. So capturing the data about the customer or understanding data about the wider market. Feeding that into how I interact with my customer or how I define my strategy.

That’s a really big piece. So at the moment, it’s like: Help me save some money with automation. Help me with services customers with a chat bot; or help me with this comply with post-trade data regulation. But the next big step will be how am I pulling all of that data together to inform which business areas I even go after us as a business unit.

So that’s the next thing that we’ll start to see in terms of actually driving changes in business strategy; but the growth is certainly there, and I normally never quite believe these forecasts from the providers, from the consultants about where things are going. But in this case, they may well be right.

David Yakobovitch

So Geoffrey, we’re kicking into high gear in 2021. We’re starting to look at the predictions for firms on AI tech strategy and data strategy. you mentioned the next 24 months we’re going to see this continue to accelerate. What do you see ahead over these next two years? 

Geoffrey Horrell 

We’ve seen a huge adoption  of NLP. It’s really coming of age now some of the new advances in the deep learning and word embedding models increase the accuracy of it, more availability of the tools, more appreciation of the use cases that it can serve and really deliver in terms of being accurate. So you’ll see NLP move front and center into the mainstream and it won’t be seen as  an alternative thing or a niche thing. It’s going to be a core capability.

There’s been a huge focus on data science platforms. On automation of model tuning and all kinds of different pipelines produced. There hasn’t been the same level of investment in the data engineering side in terms of, I want to say data engineering. I don’t mean the pipeline to the database technologies; I mean the actual data itself. So, how do we get better at cleaning the data? Cleaning the data is a wrong term. All of us need to stop saying cleaning the data. We need to break down what that means: linking the data, enriching data, identifying outliers, filtering all the different steps that are actually incredibly valuable. How do we get better tooling, better standards around how we work with that data? 

That will be the next thing. And I see a lot of investment, a lot of new startups, a lot of seed capital going into that area. I’m watching that very closely. And then, as you’ve mentioned, there’s also that piece around: Okay, so if we do all this work, if we’re really scaling this up, how are we monitoring and making sure models still work? And obviously COVID has been a big shock to that and probably the other thing that we can talk more about.

David Yakobovitch

As we think of the post pandemic world, I’ve spoken with a lot of alternative data providers and leaving funds on just data houses on Wall Street in New York City and in Connecticut.

Companies have told me, mostly off the record, of course, that: How could the model have anticipated? I started my career, Geoffrey, in actuarial science: building the models, doing the sensitivity analysis and all the testing for here. But you can only generate so many models and so many scenarios, I guess the question is moving forward: Are we going to always develop new models now that are going to take into account, whether insurance policies call it acts of God events like COVID, COVID 20, COVID 21? Do we have to recalibrate for that?

Geoffrey Horrell 

It reminds me of one of the first jobs I had was trying to sell financial statement time series data to quants. So then we got 20 years of history and then it gets nothing, that’s not enough. I need to go back to the oil crisis in the 70’s. And I’m like: Nobody has financial statement data or other data sets; also the world is so different. Why would you go and try to pick that event to back test? Because the market today is not the same as the market was back then.

So our historical approach is to go back and look at the financial crisis: .com, boom bust, oil crisis, whatever those historical events were impact test our model. The problem is that when these catastrophic events happen they’re not the same shape as the previous one. Like I was an economics student. Great. So this beautiful kind of the business cycle goes up and down in this nice way. And are we at the top or bottom? I don’t think people really believe that anymore, because it’s such an uneven world. And so the approach you think you have to take is perhaps moving to more of like a synthetic data approach.

So  how do I generate synthetic data that simulates as many different unusual scenarios as I can think of. And even beyond that, kind of use agent-based modeling or some other approach to simulate ideas that I can come up with. Because if you can think of it, you can guarantee that it’s something different that’s going to happen. But by definition, if I come up with a list of unusual events, my model isn’t going to cook with an unpredicted event, because I’ve just tried to predict all the ones I can think of. So we’ve seen the rise of people trying to create data sets in a synthetic way and use agents and other techniques to try to generate, well: what else could happen?

The benefit of those is you’re looking with today’s market infrastructure of how your model would respond to that versus going back in time, when people had 15 day holding periods, no, five minute holding period for your stock. So that is going to be an interesting thing to see how people do that and how you say: Yeah, my model is good. I’ve back-tested it on entirely synthetic data that predicted the future. I don’t think anyone’s quite ready for that, but that’s what we’re going to see. I’m excited about the importance of data scientists, data engineers that are going to be part of that. 

David Yakobovitch

That’s right. When you mentioned, Geoffrey, about synthetic data. Back to one of the key findings in your recent report is that, back in 2018, almost 30% of firms did not use any alternative data. Now in 2020 and 2021, it’s only 3% of firms who do not use alternative data. 

Geoffrey Horrell 

That’s been a massive shift, and the COVID  has amplified that because traditional data sets there’s just a massive lag in when you start to see those signals. So if I can look at footfall, if I can look at job postings, if I can look at shipment data or satellites. Data is telling me stuff about the economy or about companies or about sectors, no. Versus the government report that’s coming out or versus somebody doing a forecast that is of various reliability. I can actually get kind of like no casting from these alternative datasets, but also it’s just taken a long time for people to work out exactly how and where the apply that alternative data.

It’s still patchy. Which countries, which sectors, which places it can apply. So that’s only going to grow in use, but also you have to match that against traditional datasets. We found alternative data sets that brilliantly predict the things that the sell side already forecasts. And you’re like: Well, that’s not really helpful because we already have that data. So you have to have both together to kind of get the full picture. But again, that’s something that was alternative. Maybe it’s not alternative anymore because everybody is doing it.

David Yakobovitch

Well, we’ve only just started unpacking insights from the report on the rise of the data scientists from Refinitiv. Geoffrey, where can we learn more about all these trends and discover more of the insights?

Geoffrey Horrell 

So, you can find the report on refinitv.com, on our website. And there’s a lapse page there as well, you can see all the details of our different projects. refinitiv.com/mlreport2020. You can remember that. I’m sure you put it in the notes, but it’s a great report. We look forward to people commenting on it and giving us feedback on what they think and what they see as the future. 

David Yakobovitch

Well, I’m looking forward to the next report in 2021 to see as these trends continue to reaffirm our convictions, our thesis, and to generate more alpha for different companies in the financial sector and all sectors across the globe. Geoffrey Horrellhead of Refinitiv Labs from London, within The London Stock Exchange Group. Thanks so much for joining us on the HumAIn podcast. 

Geoffrey Horrell

Thanks David.

David Yakobovitch

Thank you for listening to this episode of the HumAIn podcast. Did the episode measure up to your thoughts and ML and AI, data science, developer tools and technical education? Share your thoughts with me at humainpodcast.com/contact. Remember to share this episode with a friend, subscribe and leave a review, and listen for more episodes of HumAIn.

Works Cited

¹Geoffrey Horrell

Companies Cited

²Refinitiv Labs