Richard Masters, vice-president of information and synthetic intelligence (AI) at Virgin Atlantic, is an skilled in enterprise information, however his profession started someplace totally different – area. Earlier than transferring into analytics, Masters accomplished a PhD in astrophysics on the College of Oxford, engaged on two main tasks.
“One was constructing an instrument within the Very Giant Telescope within the Atacama Desert in Chile, which is weighing galaxies, and the opposite one was taking a look at darkish matter in galaxies utilizing information from the Hubble House Telescope,” he says, reflecting on his tutorial experiences and time as an astrophysicist.
An outsider trying in may assume there’s little relationship between high-level science and creating customer-focused companies for main firms. Nonetheless, Masters says his work observing the grandeur of the cosmos offered the proper preparation for coping with the trivialities of enterprise data at Virgin.
“You have got loads of noisy information that is available in from these telescopes,” he says. “You spend most of your time eradicating noise so you may take a look at the galaxies and the celebrities. When you’ve eliminated that noise, you’re left with a sign that you just’re going to do some evaluation on to get your perception.”
It’s an method that echoes by means of to his current position. Masters says his longstanding give attention to tremendous particulars has develop into an obsession as an information chief: “Can I simplify and take away as a lot noise as potential from something we’re doing?”
Main information initiatives
After leaving academia, Masters utilized his expertise in a variety of industries, together with as an information evaluation guide taking a look at cyber intelligence and graph networks at Detica. He then labored as an information science supervisor with guide EY earlier than his first three-year stint at Virgin between 2018 and 2021, the place he rose to move of information science and engineering.
Masters left the airline to return to EY, the place he was given the prospect to construct merchandise for compliance and behavioural analytics as an assistant director. After two-and-a-half years honing his craft at EY, he returned to Virgin Atlantic in November 2023. He was wanting to step right into a senior information management position at an organization he knew and revered.
“I wish to construct issues, and I wish to work with individuals who perceive and belief me to construct these issues. What led me again to Virgin is that they’ve bought the fitting platforms and mindsets inside the groups”
Richard Masters, Virgin Atlantic
“I got here again due to the conversations with the manager workforce. We thought-about the alternatives the place AI could possibly be utilized throughout the sector. They weren’t simply considering, ‘We would like you to return and construct a chatbot’,” he says.
“It was extra like, ‘We all know how this expertise could possibly be used to make a number of the supply pipelines faster and the standard assurance higher, however we additionally wish to contemplate the foundational parts and make our information proper’. That target a number of areas was engaging. I got here again to Virgin to give attention to buyer behaviour and optimisation within the airline trade and to consider the issues we are able to do to fine-tune decision-making practices.”
The draw of the Virgin model was additionally a giant issue. “I wish to construct issues, and I wish to work with individuals who perceive and belief me to construct these issues,” he says. “What led me again to Virgin is that, like at EY, they’ve bought the fitting platforms and mindsets inside the groups. However the alternative to personal the end-to-end course of right here was unbelievable.”
Wanting again on his profession journey, Masters says specializing in tremendous particulars has allowed him to assist his government friends see the advantages of data-led change. “Readability of thought is important,” he says, summarising the traits that outline a profitable information chief within the age of AI.
“It’s essential distil all the things that’s occurring. Even on the strategic degree, there’s noise and opinions, nomenclature and obfuscation. Technical individuals use numerous acronyms. In any respect ranges, you wish to simplify issues as a lot as potential. Which means asking your self, ‘Do I perceive what I’m speaking about?’,” he says.
“The best way I can do that’s to supply clearer explanations. An instructional background helps with that requirement as a result of you may have expertise in distilling advanced ideas into one thing extra easy. That simplification can also be vital while you discuss to your workforce, so you may translate and say, ‘If we are able to enhance our Web Promoter Rating, for instance, we are able to present how we’re utilizing information to enhance individuals’s experiences on the airplane’.”
Managing new implementations
Masters says his duties cowl all facets of information use throughout the organisation, from analytical information engineering and upkeep by means of to the operation management centre and buyer and industrial purposes. The opposite vital component of his position is main AI developments and asking key questions on implementations.
“So, how are we utilizing AI?” he says. “What platforms ought to we use and who ought to we companion with? How will we guarantee applied sciences are efficient and that we’re investing in the fitting areas? And what does accountable AI imply for us? For me, that query comes right down to our use of information and augmentation – that human loop of assurance round information lineage.”
Masters says the importance of sturdy AI practices throughout enterprise processes continues to extend, significantly for governance and ethics.
“AI has develop into an extension of our information safety course of. We be sure that focus is vital to our work,” he says.
“Even earlier than current developments in generative AI [GenAI], bias in machine studying has at all times been a consideration while you’re constructing or utilizing fashions. You could make individuals conscious of the potential for bias and assist them perceive that, once they use these instruments, they generate extra information. Our individuals then have duty for managing that information.”
Masters says a few of his huge achievements through the previous 18 months at Virgin have been centred on data-enabled decision-making help for key enterprise areas. He says perception is used to spice up operational processes, hone buyer companies and develop new data administration capabilities.
“Our capacity to guarantee the information and fashions are appropriate has been a giant achievement for us,” he says. “I’ve bought the workforce right into a mindset of asking, ‘Is it proper?’, and constructing capabilities. We’ve created this manufacturing unit on the platform, which I’m happy with as a result of I can see the place we are able to maintain extending that method.”
Creating AI capabilities
One space of serious progress is information foundations. When Masters returned to Virgin in late 2023, the airline’s government workforce was keen to use AI.
He helped implement the Databricks platform throughout his first spell with the enterprise. After rejoining the agency, Masters explored how the expertise might present the foundations for generative AI purposes that optimise companies and quicken decision-making processes.
Virgin is now utilizing Databricks, significantly the expertise specialist’s Unity Catalog, to carry disparate information sources collectively and create a centralised location for the data that powers the agency’s AI-enabled companies. The airline is utilizing GenAI, together with Databricks’ interface Genie, for a variety of initiatives.
One instance is summarisation and categorisation. The well being and security workforce is utilizing AI to determine precedence incidents somewhat than trawling by means of a protracted record. This method is being utilized as a mannequin for sifting and prioritising points in different areas of the enterprise, together with buyer care messages through e-mail.
The newest utility of the method is for data administration. Virgin’s workers use totally different programs, equivalent to Genesys for the decision centre, Dynamics for buyer care, and apps for the crew that historically retailer information in separate locations. Utilizing Databricks, information is saved in a single location. Workers then run language fashions to go looking and summarise data.
“You possibly can preserve that one dataset as insurance policies replace after which ensure that everybody’s singing from the identical hymn sheet relating to our totally different programs,” says Masters. “So, that’s the place GenAI may also help you handle information as effectively.”
He says the important thing lesson from exploration into GenAI is {that a} resolution in a single space of the enterprise will be reused someplace else: “What we’re doing is constructing a bunch of capabilities, in order issues come up, equivalent to doing one thing with our app relating to GenAI, we are able to draw on these capabilities as we’d like them afterward.”
Delivering seamless experiences
Masters recognises that the endpoint for data-enabled journey through the subsequent few years is extra personalisation and higher-quality, seamless experiences. Nonetheless, reaching that time means answering some difficult questions.
“How a lot do you wish to learn about AI as a buyer?” he says. “The reply is, you don’t, you simply need your expertise to be higher. So, clients need us to know extra about them inside the fitting context – to cope with issues shortly within the occasion of one thing occurring or present higher data alongside their journey.”
We wish to transfer to a degree the place you shouldn’t must verify in [for your flight]. We must always know sufficient about you that we simply want to make sure you flip up on the day of your flight. Every little thing else is dealt with by means of identification and the system Richard Masters, Virgin Atlantic
Masters needs to make use of AI, digital experiences and information to enhance the airline’s professionals. Virgin’s workers ought to be capable to mix these technological capabilities with their very own experiences to allow higher companies with out AI being a part of the dialog with the client. Throughout the broader trade, he notes the shift to AI-enabled automation.
“We wish to transfer to a degree the place you shouldn’t must verify in [for your flight],” he says. “We must always know sufficient about you that we simply want to make sure you flip up on the day of your flight. Every little thing else is dealt with by means of identification and the system. That’s the place the trade is transferring. And truly, AI may also help us within the transition to that state, versus constructing huge options that may develop into out of date or prohibit us from transferring shortly.”
Masters says reaching that stage would require trade alignment on frameworks, terminologies and instruments, significantly relating to integrations with different parts of the passenger journey, equivalent to connecting transportation and vacation spot resorts. Nonetheless, he expects progress to be speedy because the promise of rising expertise is realised.
“You’ll see extra alignment throughout the trade through the subsequent few years,” he says. “At Virgin Atlantic, that progress goes to be round the place we are able to use information to let our groups unleash extra of what makes our airline particular.”
