Leader Bio - Elliot Banks

Elliot Banks, Chief Product Officer of data-driven fintech start-up BMLL Technologies, discusses how the pursuit of knowledge in new and unexplored fields shaped his journey through education and work, and led to him entering the then-emerging world of Data Science.

The Unknown

Computer science is the most recent and radical paradigm shift in human knowledge. Software Developers, DevOps Engineers and Data Scientists work at the bleeding edge of our acquired technological development, in a field that constantly drives its own evolution.

This perhaps explains learning’s central place in Engineering culture. Not only does modern technology sit on the shoulders of millennia of accumulated understanding, but it changes constantly and rapidly. Technology attracts those with a deep curiosity for the latest developments, and hinders those who lack this drive.

A passing glance at Elliot Banks’s CV tells you that he learns quickly. He holds a First Class Physics BA and a Distinction in Maths from Cambridge University, as well as a Theoretical and Mathematical Physics PhD from Imperial College, London. His career, however, demonstrates that it is an appetite for the new and unknown, as much as the pace at which we learn, that shapes a person's path and their ability to lead.

BMLL

Elliot is Chief Product Officer at BMLL Technologies, a financial data and analytics platform provider which takes granular, historical data sets from different sources, often in disparate, inaccessible forms, and standardises them. On top of this is built a suite of analytics and visualisation tools enabling analysts, quants and other users to derive greater insights from the data and attempt to predict future price movements.

Customers include stock exchanges seeking to understand their market share and competition; execution services such as investment banks and brokerages; ETF (exchange-traded fund) issuers looking to understand how their products compare to competitors; and smaller hedge funds. Many larger hedge funds have in-house teams that build similar capabilities to what BMLL provides, but smaller competitors or spin-outs of these, while knowing what good looks like, often lack the resources to do so themselves.

Elliot became BMLL’s CPO in mid-2019, aged just 28. He joined two and a half years earlier as a Data Scientist, a role he compares to that of a Sales Engineer. He was involved with the building of the product, but also in presenting its potential use cases to customers. This multifaceted position required not just excellent technical ability and understanding, but also the ability to think like a customer, and talk to them in their language.

Elliot feels he learned these skills during the year he spent working in private equity for Macquarie following the completion of his Masters.

“I would massively advocate doing that, even for people on the technical side. Having that exposure to a more commercial setting was immensely valuable to me.

“I only did it for a year, but the broadening of my skillset was really valuable in a way that I certainly think has helped me. Certainly on the softer skills, on the presentation and communication side.”

A Learning Machine

The suggestion that, on completing his undergraduate, he could have predicted becoming a Data Scientist or a CPO prompts a laugh.

“I didn’t know what it was then, to be honest. Even when I was offered the CPO role, my wife said ‘That’s not a real job.’ I pointed out that Facebook had one, and that made her a little more comfortable”

Elliot learned C and C++ during his undergrad, but learning in a theoretical environment only took him so far.

“I'm a very practical learner, which is quite odd to say of someone with a theoretical physics PhD. With coding, I really need a practical application to work on.

“Doing something in C or C++ as the first language I'd ever touched was difficult. You get introduced to a lot of concepts. A lot of object-oriented programming is quite abstract, and it's only when you see something at a production grade level that you realise, ‘Oh, that's why I need a class. Oh, that's why I need to use these factories and these design patterns.’ So it's very difficult at university to really learn it.”

It was during his PhD that Elliot feels he really became a proficient coder, as well as learning valuable energy-allocation skills that would serve him well in the start-up world.

“Doing a PhD, you start off on something and you genuinely don’t know if you’re going to go anywhere. You might spend a month or two doing something, getting absolutely nowhere. Then at some point, you have to call it.

“Dev work, especially greenfield dev work, is similar, but obviously the timeframes are shorter. You try something for a week and at that point say ‘Right, let’s back out and let’s reframe the problem.’

“That ability to be able to persevere, but also to know when to quit, was a really useful skill I got in my PhD. And that really helped me on the development side of things.”

Elliot’s PhD focused on the behaviour of 5-dimensional black holes, and explored the implications of their mathematics for real-world phenomena, such as the superconductivity of certain materials.

Following it, he naturally turned his attention to a then-emerging field that also sought to apply predictive mathematics to real-world problems: Data Science. The term had been coined shortly after he graduated his Masters, when it appeared in a Harvard Business School article in October 2012 titled Data Scientist: The Sexiest Job of the 21st Century.

“Before then, you had lots of data analysts. You had people using data, but data science itself wasn’t a thing.”

Following his PhD, he attended a two month data science fellowship hosted by ASI (now Faculty). This, he says, was an opportunity to gain exposure to commercial data science, another experience he feels benefits his career in a start-up.

“In a start-up, you can’t be too theoretical about things. You have to be pretty practical, pretty grounded. There’s no point building the perfect solution if it’s going to take you three years to get there. We have to get things that turn commercially viable in a time frame that works for the business, very much that ‘fail fast, learn and iterate quickly’ mentality.

“That's the sort of thing, with some of the younger members of my team, I always try and ground them in because it's very easy when you come out of university, you've got a good degree. You've learned about how some machine learning models work. You learn how to do some data science. It's very easy to get swept up in doing something cool and interesting with the data, rather than answering a practical question. And ultimately data science is valuable when it's answering a real legitimate problem and solving a business case.

“Then, you are using data to help inform the decision rather than doing machine learning for the sake of machine learning.”

Explorative problem-solving

BMLL’s platform wrangles petabytes of data, and has a very open-ended range of potential applications. This, as far as Elliot is concerned, is both the challenge and the attraction.

Elliot now runs a team of data scientists, tasked with expanding the potential use cases of the product, communicating these to customers, and responding to their requests. Every facet of this requires an open, inquisitive mindset.

“There’s an awful lot of smart people in finance, especially in the electronic trading side of the industry. Often a problem gets presented by a customer as a solution. So the customer says, ‘I want this.’ And they don’t actually mean that, what they mean is, ‘I’ve got this problem, and this is how I think you should solve it.’

“Now that isn’t necessarily the solution every other customer wants. It’s not necessarily the solution they want. It’s the solution they’ve come up with. That’s a hard thing when we have very sophisticated customers who understand the technology, and have their own viewpoints on how a solution should be built.”

Elliot ensures his team are enabled to solve this problem firstly through empowerment, fostering a culture whereby everyone is encouraged to think creatively and challenge ideas, and secondly by closing the gap between Technology and Product.

“We don’t feel there’s a business team and a separate technology team. We’re one team.

“We’ll talk to the team leads, we’ll talk to the devs, we’ll talk to the CTO. They might say, ‘Well hang on, they’ve asked for that, but why do they want that?’ And we have a culture where it’s like ‘Okay, here’s the problem. Let’s bring some developers in, let’s bring the product team in together and come up with the best solution.”

This brings about a sense of shared ownership, trust and transparency, that invigorates this creative, explorative problem-solving process.

“We don't just say, ‘Here's the requirement. Go and do it.’ We say, ‘Okay, this is the context of what we're trying to do. This is why we’re trying to do it. Here is what we think the requirements are, but what do you think?’ We encourage people to give us their opinion.”

Managing without a handbook

This creative, collaborative approach to problem-solving is typically embraced readily by the BMLL team, and by new hires. It’s not something the company’s recruitment process specifically targets, though Elliot admits there may well be an element of self-selection about the kind of mindsets that are attracted to working there.

It has, however, posed a significant management challenge in the face of COVID-19.

“There isn’t really a handbook on how to manage teams during a global pandemic,” says Elliot. The challenge is balancing the creative and collaborative demands of building software with the productivity and flexibility that allowing people to work from home offers.

He feels the company has settled on a comfortable hybrid model, where most staff tend to come into the new office two or three days per week.

“It’s taken us a while to land in the right spot, but it feels like we’re pretty good there now. We’ve just moved into a new office, it’s got a really nice vibe. We’ve landed on that quite nicely.”

Flexibility and a light touch have been some of the most important managerial lessons Elliot has learned.

“The biggest thing for me is to empower your team. Give them the freedom to succeed, give them the freedom to make decisions. It can be a really scary thing at first, but that empowerment is really important.

“You give people the freedom to come up with ideas, while knowing when to step in and provide a little bit of guidance, a little bit of leadership. And when to actually sit back and say ‘You’ve got this.’”

The trajectory, he says, generally goes from developers beginning their careers being told what to build, then gradually having greater ownership over their products as they progress into senior and lead positions.

“Over time, you become the best developer and you know more of the system than other people. Then people are like, ‘Go and lead this.’ There’s a natural temptation to say ‘Okay, well, I’m just going to dive into that code base and do it.’

“But actually, that’s the opposite of what a good manager and a good leader should do. They should let their team figure it out, manage that and educate them. If you’re  good enough, the company runs with you, which is the ultimate thing. And you can sit there and steer it from high level.”

Next steps

Start-up life is, in many respects, harder to predict than the behaviour of a black hole, and no data scientist claims to know the future. In terms of his and BMLL’s next steps, however, Elliot sees a couple of current trends continuing.

The first is to expand on BMLL’s current product offering. As is the way with start-ups, many of BMLL’s current revenue streams weren’t predicted in the company’s early days. Elliot alludes to Google, who initially sold their search engine to companies as a means of indexing their own systems, and only later on realised the potential of using it for public search.

Google’s diversification is, coincidentally, one driver of another shift that Elliot and the team are predicting, namely creating a multi-cloud strategy.

“Everyone at the moment feels like they’re picking their cloud dance partners. For us, that means long term we’ll almost certainly be led down more of a multi-cloud strategy by customers saying ‘This is great, but I need to be able to deploy this on my Google Cloud stack’, for example.”

BMLL’s expanded product offering and increased technical complexity will see Elliot’s own managerial arc continue to grow as the company and his team expand. Unsurprisingly, learning is high on the agenda.

“I was fortunate to be in this position relatively early, but it still means there’s an awful lot of learning on my side. And also a lot of challenges that we’ll face as we start to scale out. That’s what I’m really excited about, from a personal development point of view.

“It’s all about the scale of the data and analytics. We’ve got petabytes of data, and that is a real challenge.”

Using data in an impactful way will always be closely tied to Elliot’s notion of doing your best work.

"It means doing work that has an immediate, visible impact. The worst thing you can do in work is spend time on projects that don't go anywhere and don't make a difference for the business or customers.

“Throughout my career, I've always looked for roles and opportunities where I can make that impact, whether it's in a commercial, academic or technological sense, and look to empower my teams to do the same. And the culture that we've built at BMLL really helps develop that.”