Seventh interview with interesting people within Fintech, this time Sean Bowen of Push Technology...
Sean drives the spirit of innovation at Push Technology with his market vision, business acumen, technical expertise, and unbounded energy. With his keen understanding of the requirements for real time data streaming technology to power mission-critical business applications, he founded and has invested his talent in leading Push Technology to serve the real time data delivery needs of markets worldwide. An entrepreneur at heart, Sean founded his first company, RFT Solutions, to develop business-essential, real time dashboards for operational banking systems. Prior to that he headed programs for development of bespoke financial reconciliation systems at Bond Technologies (now NorthPoint Solutions), a consultancy with domain knowledge in hedge fund operations and the technology applied to the tracking, reporting, valuation, and risk of complex portfolios for companies with AUM in excess of $1B. Sean’s programs comprised large, high-tech, consulting projects in the financial services sector for multinational clients including Deutsche Bank and Goldman Sachs. Earlier, Sean managed software development for ICMA (International Capital Market Association formerly the International Securities Market Association), which is internationally recognized as the preeminent business school for financial markets; and, he worked as a business analyst at the Chartered Institute of Public Finance and Accountancy that offers a range of high-quality advisory, information, and consultancy services to public sector organizations. Prior to launching his high-tech career, Sean was a member of the Royal Air Force where he worked at Strike Command headquarters. Sean is a frequent and sought-after speaker at industry events.
(Some discussion about my Fixed Income work and "the list" lead to recollections about ABN Amro, where we join the conversation...)
I worked in Fixed Income myself for years, going way back to ABN Amro, doing trade support. In fact working with my old 1st team rugby captain.
Was that at 250 Bishopsgate?
I had two stints there, I helped with the migration from the old office to 250 Bishopsgate. Second time was when the firm had already moved.
Who are you?
Sean Bowen. I left the Royal Air Force around 2000, I wanted to get into IT when I left. The first financial job I had was for ISMA. I was fortunate enough to get into a consultancy firm early and get thrown into the deep end.
Going back a little, I have a big sporting background, love challenges, hate losing. Consultancy was great as I was thrown big challenges regularly. I got dropped into a deep end on a number of occasions. I never liked to lose when given a challenge so I was successful in the projects I was given. ABN Amro was a great example, we got parachuted in there because they could not migrate the data internally prior to moving their offices to 250 Bishopsgate. We said we could do it, but it had to be in Excel because we only had two weeks. We managed to do the job in the two weeks and our client looked like a superstar at the time! It was a great project, great fun. Probably my first big challenge in the banking sector. That reconciliation project led to me and a colleague to build a VB6 reconciliation application that replaced their internal Sungard system. I became something of a reconciliation expert, I worked for Rabobank for a while on their big flagship STP program. I lived in Utrecht for a year doing that. And then, like all good consultancy practices, I came back to London and I was thrown into a completely different field. I was working on bond trading with ION Marketview, throw a bit of treasury in there as well for good measure. I managed to get quite a diverse set of skills from the financial markets, working with bonds, equities, front office and back office.
While I was working with that consultancy firm I saw a bit of a gap in the market, probably two years too early. I was looking at how banks were tracking their data in real-time. When a bank does a big trade, it’s often with a bunch of counterparties and they were not getting the visibility in real-time on what trades were actually settled. So they were not able to put their money in the bank on an overnight deposit. I thought that was a bit of a gap and when I dug deeper into the topic I saw that banks were breaching bought stock and house limits. Putting real-time visibility onto that in the banking sector was the plan. But that was two years before the markets crashed and so two years before everyone was focusing on implementing it.
It’s worth stating, that was what got me into examining this whole space. I had connected the back-end systems, trading systems and risk systems; but, we were still giving the data to the Executives using a web-page that was not very dynamic and not very scalable. As a result, I turned to the internet to see how we could solve the real-time, scalable data challenge for dashboards. Also for other markets, I was looking at other technologies around 2006. or example, in eGaming I saw that in-play betting was happening and I thought how are they going to make that work when they are not used to the throughput, scale and performance. It was a good time to launch Push Technology and solve the last mile piece. How do you deliver massive scale and massive throughput over the internet and still assure low-latency so people can have the most compelling experience.
Add into that I have played Rugby and Football for fun, that gives you some insight into the person I am.
What firm do you work for?
I am the founder and CEO of Push Technology. We have a flagship product called Diffusion. Diffusion is an intelligent event-data platform which most everyone in eGaming knows as we have been with most of the big players for a number of years. We have transformed the whole market with in-play and how they do it. We have had some strong success in the financial markets as well, that has really blossomed in the last couple of years. Everything was compliance and risk for quite a few years, nothing was changing in the big financial houses and all the FinTech firms were nipping at their heels. Now they are out of that cycle and are seeking to transform what they are doing and to defend their market shares. As you know, there are a lot of systems that have been in banks for many, many years and they have not really changed. These systems require upgrading and extending out to get better access to their customer base.
Trading is a great example, I am sure you are aware of single-dealer ,multi-asset trading platforms. They were built primarily for corporations and banks to trade among themselves; but, now that’s getting opened up to the public in fact, looking at Consorsbank, they were one of the first firms to approach us to say that they wanted to get trade prices into their retail banking customers hands. And that’s happening more and more in the banking sector. It is no longer a matter of can you stream prices to a couple of thousand traders. Now, it is can you stream prices to a couple of hundred thousand people? Push Technology has been primed for that for years, we started out in the gaming space which was to me very similar to financial markets. Betting opportunities instead of trading opportunities where you are betting on sport rather than trading financial instruments, but you have prices that are changing in real-time and traders that want to trade on those prices.
Traders are trading much greater volumes – instead of a couple of hundred or a couple of thousand traders we were dealing with tens of thousands or hundreds of thousands of users. Sometimes even millions! The one difference is that the betting market is not as volatile as financial markets but it has its own scaling challenges that we had to meet – many people wanting to trade in real time. That was a great test of our product early on. Throw in that we have been pulled into a number of banks and interdealer brokers – we had three of the top five for a while before two of them merged.
We have been stretched with trading systems in banks with throughput and we have been stretched with scale of connections in gaming companies. Both have been excellent tests of our platform and really helped us to build something that now I am starting to see other markets want. They did not know it was possible.
We looked at the internet as the bottleneck and we did not follow traditional approaches to moving data. We had realised traditional approaches were hugely inefficient and we actually came to market with something unique and that has given us a competitive edge from day one. It’s quite funny that the key feature we are talking about is how everyone else moves messages like a Postman – they always deliver every message they are given, they never look inside the message, they just deliver. That’s a really inefficient way of moving data, especially over the internet, since quite often you don’t need to move the entire letter, you just need to move what has changed. We got inspiration early on from video and that is how we delivered our data platform. The way video works is that the very first frame you get has every pixel and the second frame is not every pixel, it’s only the delta, it’s just the pixels that need changing to convert the one frame into the next. We built our platform around that as a key concept. In the gaming space we found we were driving down the bandwidth required by 99% and that’s why we transformed the whole market, we were enabling customers to go from 20 betting opportunities per game of sport to 150 which was the entire market that was available then. We were able to deliver prices so quickly because we were very efficient and we weren’t flooding their network This meant we were able to handle point-by-point betting on a game of tennis which was impossible before.
Rolling forward into the banking sector, it’s worth saying that the banks were fairly efficient with how they moved data. However, we still brought huge efficiency to it but what we have done is realise that there are a number of products going around such as Kafka that is phenomenal, everyone is adopting it. But it delivers a firehose that none of these systems were built to accommodate. We are finding outages and people are getting data they should not get. As you know, within financial markets not every piece of data can go everywhere. You certainly don’t want customers seeing other people’s data. There are certain instruments you cannot trade in certain territories, so having an ability to control that firehose and route data where you need it to go and stop it going where it should not go is becoming hugely important for us as well.
The product has really evolved over the years from the original highly performant, highly scalable piece of technology that hugely helped the gaming market to scale and financial markets to deliver trading systems over the internet to something you can pick up and integrate with back-end systems quickly and easily. Pressure on companies is building every year, to deliver more with less budget. Being able to hook into back-end systems that, in some cases, have not changed in years is one concern, another is that new systems are so fast old systems cannot keep up with them.
How do you handle that throughput? By being able to put something that sits in-between, that gets data from these legacy platforms and controls the flow of data to highly performant messaging systems and enable people to change that data in-flight to meet the demands of what a client wants to consume. That’s a challenge we see quite often, back-end systems – if you want to change something in a Bank you have to wait three months, with Kafka people don’t want to change it as it rebalances the whole platform. You don’t really want to touch these back-end systems, but the data they are moving never matches what your client wants to consume. Being able to do data wrangling on the data in-flight and change it into a meaningful format that is ready for your applications and customers to consume is, we found, a very compelling part of what we have built.
It’s amazing how, early on, everyone was all about delivery and now it’s all about data-wrangling and control of the flow of data.
What do you do for Push?
Ah, interesting! Probably everything if you speak to half of the company. I have my fingers everywhere, whether it’s sales, marketing, finance, product. I think it’s important as a CEO to know every facet of the business, it’s very hard to lead if you don’t. I’m fortunate that I have been exposed to multiple parts of different industries so it has given me the tools to get involved and I also think it’s important when you are building a product and selling it into markets you have an overall view, not just a one-sided view of “I want to sell more” but “How do I sell more? What is needed in the product? How do I support a customer? What kind of service do I want to deliver to a customer?”.
I get my hands dirty half the week and the rest of the week I am managing, leading and performing my CEO duties. It’s not a quiet week and it certainly isn’t 9-to-5.
What made you start Push?
I have had so many people say that to be a founder of a start-up is a huge risk. At the time it was, people said you must be brave to do it. I said no, probably stupid! I just didn’t think about it to be honest. I saw a gap in the market, a challenge I wanted to fix and I knew I could do it. And that’s what made me start the company. I saw a statistic that 1% of companies last ten years, so we are doing something right. In fact we are probably having our best ever year, right in the middle of a pandemic! It has been an interesting journey and now I think I would find it hard to work for someone else, I like the freedom of being the CEO and making the big decisions. I know that not everyone is comfortable with making difficult decisions but I quite enjoy it. I think I would be bored as well, working for someone else. It’s a million miles an hour at Push, everyone is on the same journey, we have built a fantastic culture and it’s an exciting place to work. I think working anywhere else would be difficult now.
What has your personal journey been?
I started playing semi-professional football and I had an injury. Everyone thought I was going to be a professional footballer. I come from the wrong side of the street in London, from Woolwich and Plumstead. Not an affluent are by any means and I had to fight my way out of there. Soccer felt like the way to go at the time. But I got injured and realised my career could be over at any time. I didn’t want to go back to school and study for A levels so I thought what could I do that would give me a similar level of training, knowledge and a leg-up in life while also experiencing life. And that’s what made me join the Royal Air Force. I think that what really made me do it was my parents telling me that there was no way I would ever do it! As you have probably worked out about me, I hate being told no! That was it, the point made, I am joining up. Unfortunately they civilianised the role, my role was switched to being a civilian and I didn’t want to go into another part of the RAF so I came out and got back into civilian life. It was a phenomenal experience for me, certainly changed my personality and the way I looked at life. The first time my parents met me after my Passing Out parade (the parade at the end of the basic training period) they said “where has our son gone? You have got manners!”
I would recommend it to any youngster that wants to grow-up in life, it’s a real shock to the system going through it and I certainly came out of it a lot better than I went in.
What do you see going on in the world right now where people are gaining real commercial advantage from the products of your firm?
I don’t think it’s just my product in my firm. Let’s cover the big question which is affecting everyone - the pandemic. Working within banking you see that when there is volatility, it can be good and bad. This has been terrible for a lot of companies but has also been great for other companies that are remote working and offer other creative experiences that are more relevant to the world today. The pandemic has opened up opportunities for people. It has not been massively positive or negative for us yet. But, long term, I think it will be hugely positive for us because, as I mentioned earlier, we were focussing on how we could deliver data efficiently over the edge and deliver great performance at huge scale. With the pandemic everyone was sent home to work and so our office is now in our house with our dogs and kids going mad and that’s the new normal, rather than going into a building in London or wherever and performing my daily duties there. When I was in my office I had a dedicated network, with all my systems around me. Of course, everyone else was experiencing the same kind of work life. Now we are all working from home, on a shared network, with all of these systems that were just not built to be delivered over the internet at the scale and pressure the network is under today.
There were some crazy statistics from the likes of Zoom about the adoption of video calls that would have absolutely hammered the network of all of these telcos and no surprise that there were a bunch of outages early on. It has led people to look at these systems now and see how they can re-architect them to work remotely. We are at the beginning of that journey at Push, we see some of these changes happening and we also see a bunch of companies really struggling as they cannot afford to make those changes yet.
As time goes on it will get better and better for us, but I definitely think the whole work-from-home life is going to have a huge impact on our business. We are seeing it in other industries as well, we are obviously strong in gaming and finance but we are now seeing firms in other industries approach us to adopt our technology. And not just on-premise, cloud is getting a bigger push, pardon the pun. Not being in the gaming and finance space where people want these systems as close as possible to the source of data for latency requirements. When you get into other sectors latency is not so important, reliability and scalability and delivering data within time is important, just not at low-latency speeds. They are more than happy to go to the cloud so it’s a very different adoption model when you are trying to sell to those markets, very different product required, very different sales cycle required. It’s been interesting as this shift has happened in the world, what we have learned from it, what opportunities we have discovered from it.
As the years go on, this is not going to go away anytime soon, we might have a vaccine now, but we know these viruses mutate and this will change the way we work forever. And I think that’s going to be positive for Push in a big way, just because of that whole data efficiency, scalability and performance play we have had for many years over the internet.
Blockchain, distributed ledger – is that something you do?
We have had some people from the blockchain world come to us, wanting to use our technology, not for execution, more for streaming live prices. We have seen it a little, it might be one of those technology trends that is a great idea and will get traction in certain areas but not everywhere. Not to downplay it, it’s hugely valuable where people are using it.
The big one we are now seeing is the whole IoT market is starting to open up.
I have done some work with MQTT, interesting world?
It’s funny, we have been working with IBM for years, they had massive success with MQ Series. They are hugely influential within MQTT and we have relationships with a number of their architects. They would always tell us that MQTT is hugely efficient, but we would say it’s not. We’d show them ours – we have just launched an MQTT plugin for Diffusion, so you don’t need to use our SDKs, just go straight to an MQTT device without any broker in-between. It’s great, it makes it really easy for anyone to pick up Diffusion if they don’t know it and if they have an MQTT device available. Whether that’s a mobile device or a sensor. We have our own SDK connected on our demonstrations internally, we built some metrics collectors years ago. We ran a collector to see how efficient Diffusion is versus MQTT and I think it was still over 80% more data efficient.
The thing about MQTT is that the way it’s structured is very inefficient – the variable byte integer – it’s very inefficient. If you have an integer in your system you have to munge it down into the minimum number of bytes, pack it in, send it over the wire and then unpack it at the other end. It feels like a case study of Amdahl’s Law about premature optimisation. You have less data going down the wire but more compute required at the endpoints to handle the data. And one of the core points of IoT is that you have resource-constrained endpoints with a minimum of compute and memory. The broker has compute, so why do compute on the edge devices?
We have put most of our important functionality on the server. We do have the functionality for a client to send a delta message inbound which is unique to us. Again, we were looking at these sensors and thinking why would I want to send the whole payload again? It’s quite often a JSON payload, so when you have sent the names once you don’t need to send them again, just send the values that have changed. It all just seems very inefficient, so we put this on the client side and we have seen great efficiency compared to MQTT. People don’t want to get tied-in, which I understand, they want open-standards, open-source and be able to move freely. So I get the reason for adopting MQTT, but at the same time it is going to hinder your performance and scalability when you are trying to deal with mass scale and you want low latency.
Even if you use MQTT people still use things like Screwdriver inside it – to me MQTT is more like a session protocol with <your payload here>! And so you have to use something else to describe your payload…
It was funny, I cannot share the name, but we sat down once with a firm and they said “if we just use MQTT do we get all those cool features you have?
We replied – “No! You are going to lose some of them. The ones on the client side you lose, since MQTT does not have it. You get the benefit of the plain MQTT server. Deltas – gone, reconnect – MQTT is not very good at that…”
It’s interesting as you compare the two, but our job is to give options to the client, they can go down the MQTT route. I absolutely understand why they may want to go with open standards. But if they want the higher performing, much more data efficient, intelligent movement of data they can use our SDKs.
Cloud - in use with your firm? Hype or real?
We have an on-prem and cloud solution. We are not a multi-tenant cloud, we do use Docker and Kubernetes and get around it that way. We have a cloud service that is hugely efficient compared to everyone out there, probably more performant. We are still building-up some of the services round the outside, but no-one out there has this data wrangling service on the cloud. It’s a great tool and we are eating our own dogfood. We are using it right now for a skunk works project internally that will be announced at some point. I think its hugely valuable for any company to use their own products and services. We are quite excited about the lessons learned and the new product we are creating.
Cloud is great for people who want to get to market quickly and without huge expenses. It’s not cheap, that’s a misnomer. As soon as you scale-up it becomes expensive. I don’t think it’s as cheap as cloud people portray! But it is a great way to get a product to market cheaply and quickly where you don’t have a big upfront cost.
Where do you see technology impacting in the next five years on:
- Your firm
- Wider market
Me personally? I have two very active kids, both very sporty. I have been working on the whole sports sensor thing, working with a firm, embedding our technology into theirs. Just a personal little project that I am enjoying. Seeing real-time statistics on athletes as they are playing. It’s a great thing for colleges over here. Think of COVID, it’s a big driver for colleges here to get the best athletes in and previously their only way to get students in was to watch them play or get a video. So I am loving some of the developments from our partner companies working with our technologies. Since I can see how that will impact my children.
When I think about systems and solutions we use at home I get frustrated when they are not real-time. Or when there is lag or performance degradation. I am hoping more and more people adopt our product. Not just selfishly in that I want the company to do well, just that it’s a better experience for people. Everyone should have the same experience at home that they used to have in the office. The more and more our product gets developed outside of these low-latency type systems the more we will have that kind of experience.
With regard to Push Technology, we are getting pulled into all sorts of markets, all different use-cases. And it’s hugely fun to be involved. We are very fortunate that the technology stack has changed, everything used to be REST based but with Kafka, Microservices and IoT everything seems to be going event-driven. And systems are now being re-architected to take advantage of that and that’s a huge opportunity for our company and we are seeing that in some of the deals we are closing. As I said, in the midst of the pandemic it’s our best year yet.
On a bigger scale, there are so many cool technologies around. One thing I love about technology is that you do not progress at a steady pace, as technology gets better it just accelerates faster and faster. We have Quantum Computing coming around the corner next, which we are also starting to get involved with as well. You mentioned Blockchain, IoT, AI, Machine Learning… I could keep going, there are some fantastic new technologies that are coming around and changing the world and how we live. Just with this pandemic, think about the data that is now available in real-time and how quickly vaccines have been put together. Think of how long that would have taken ten years ago…
There is stuff that we don’t even know about today that will appear in the blink of an eye in five or ten years that will move things forward massively again. It’s a massively exciting time for technology. Products such as ours that can move data quickly, products like AI and Machine Learning that can analyse data quickly, are just going to accelerate a lot of possibilities for people.
Fintech – is that how you describe your firm?
There’s no doubt that we have a huge influence on financial services so you could say that. We are in with a bunch of trading platforms, back-end services, we have just got into wealth management with a huge bank. But this is bigger than finance. Like I said, we are getting pulled into all sorts of industries: shipping, retail, logistics, satellites…
It does have a big play in that fintech space and I find now a lot of fintechs are looking at this kind of technology and a number of banks have been looking at this for several years as a way of driving change much quicker in these organisations. But I would not want to pigeonhole us just as a fintech company, even though we have just won one award in fintech and are shortlisted for another! Other people may look at us as a fintech…
You think it [Push] is bigger than that?
Massively scalable real-time event data is applicable to every industry. When I think about fintech, half of that word is financial services, the other half is technology.
Go crazy – make some wild predictions about machine learning, AI, DLT, cloud, whatever!
Cloud – the acceleration of people who want market data in the cloud. I still remember from years back, you would never dream of putting market data in the cloud, it would be inside the bank, controlled..
Exactly, there’s a name from the past! We have been doing some great work with a company called BCCG, we did a webinar recently on how you can have market data in the cloud and take away some problems.
There is so much change coming so quickly. One you missed – AR. And VR. Another huge market that is trickling along that could do with a dose of Diffusion in the middle of it to make it all perform quickly.
With regard to AI and ML, a lot of jobs are fortunately and unfortunately getting replaced by technology. This has massively driven down the costs of building and providing a service. And you can deliver a superior service. You still need people to make sure these processes are working. I think it will be interesting to see how the jobs market changes over the next few years. What sort of jobs are going to be the norm? One thing I say to my children is “Security”. One thing that will never go away if you are looking at computers will be “Security”. With a lot of jobs being automated the changes to the jobs market will be very interesting.
IoT will accelerate – we’ve not really seen any of it yet. I remember all of the noise and right now, you can have a sensor in your fridge…
That’s not a surprise, people needed to build the devices first. I have seen how people are hooking them up, using technologies like REST, and I have seen how poor a lot of the services are. I love the fact that we are starting to get dragged into IoT and we will speed up that whole market as well. It will be interesting to see how that changes. I have been speaking to people about autonomous vehicles and connected car. All of them have been built for a few cars, not to scale. That will change and we will see a lot more IoT working at mass scale using platforms like ours.
Speaking to an expert in the industry we asked “when can autonomous vehicles come?” and he replied that “they can absolutely come today, but the problem of scale has not been solved”.
The second paragraph (biography) was supplied by Push Technology as was the picture of Sean. This interview was conducted remotely over a Zoom call and transcribed by Alignment Systems. No payment was asked for or received either way.
Post a Comment