How is your organisation using data and analytics to support the corporate vision and purpose?
At QuantumBlack, we see data and analytics as a core driver for continuous improvement in what we do and how we do it: how can we iterate faster, more consistently, and safely, to squeeze maximum value from our initiatives, whether internal R&D or client transformations.
We interpret data in the broadest sense, not just structured in tables, harvesting and analysing a wide variety to metric our ways of working and constantly identify inefficiencies and opportunity for improvement. Our clients adopt similar philosophies, using data to drive strategic high-value initiatives. And this “learning machine through machine learning” mindset is what we try to embed within the executive and business teams as a greater ambition than specific data initiatives.
2020 was a year like no other - how did it impact on your planned activities and what unplanned ones did you have to introduce?
Implementation times were accelerated given the global situation – projects that could have taken up to two years were rolled out in 12 weeks. We also found new ways of working with our clients remotely.
We were fortunate that some of our initiatives rolled out prior to 2020 really enabled this acceleration, supporting our teams and clients; for example, our codification of our end-to-end analytics at scale methodology (QB Protocol), ensuring there was consistency and speed in delivery. Equally, our open-source tools, Kedro and CausalNex, allowed not just us and our clients access, but anyone to build data pipelines, analyse datasets and build models.
We also pivoted in 2020 to focus much more on R&D and increased global collaboration. We were able to launch an internal technical knowledge repository, which allows teams and clients to contribute, discover and improve well tested and verified code and other technical assets to accelerate projects.
Another area we really invested time in was our own QuantumBlack community and keeping our teams connected. We spent considerable time creating ways to engage with our teams, from comedy evenings to escape rooms. Given the situation, connectivity with everyone was super important so we communicated with weekly get-togethers, providing updates on everything happening.
Looking forward to 2021, what are your expectations for data and analytics within your organisation?
For our clients, we will continue to see digital acceleration and a drive to improve customer services and journeys. The global situation has highlighted the need to have more effective digital strategies and data and analytics will be critical to enabling improved personalisation and customer experience. More of our work recently has focused in this space, and I expect it to continue to ramp up.
We will also see additional explainability and transparency requirements as further use of data and machine learning proliferates. This increase will require more scrutiny around the use of data and the risks around fairness and unconscious bias – a nascent area we have already invested in heavily and released research papers (on explainable techniques) and a model risk governance framework.
Is data for good part of your personal or business agenda for 2021? If so, what form will it take?
We are committing considerable resources to technical solutions and processes to help our teams and clients build models that are governed-by-design. We have a significant responsibility to ensure appropriate use of data and monitor for any conscious or unconscious discrimination, especially in today’s world.
As a company we also apply our knowledge of data and analytics for a number of pro bono initiatives, such as supporting efforts on climate change. Finally, I’m on a personal mission to bring more sustainability into our delivery teams and this year plan to introduce carbon footprint tracking applications to see where we can reduce unnecessary behaviours or habits.
What has been your path to power?
I have not had a typical path to where I am today. I am a self-taught coder, having always had an interest in technology: from my first personal ZX Spectrum to developing programmes on BBC Micros and building my own PCs (spending way too much at Overclockers).
After university, I went to China for a year to continue studying martial arts where my teacher asked if I could help him build an online Kung Fu encyclopedia. A friend and I learnt how to build websites and this started my journey into real technology.
After coming back to the UK, I qualified as a freelance web designer/developer and started work in this field. Since then, I have moved from helping clients re-engineer global processes supported by large technology implementation, through to building large-scale data architectures and MI solutions.
Nine years ago, I was introduced to Hadoop and became excited about the opportunities of larger scale data processing and the possible value it could unlock. This prompted me to want to explore analytics and value creation, so I moved to QuantumBlack, a McKinsey company, to build my knowledge of both data and analytics and help develop a capability to optimise building data foundations to support AI at scale.
What is the proudest achievement of your career to date?
I am most proud of the global team I have built at QuantumBlack. When I joined there were three data engineers in London, and now we have more than 100 globally. Helping to define our value proposition, shaping our technology strategy, building recruitment processes, bringing on fantastic, talented people, setting up a capability in India, and getting a chance to work with engineers far smarter than me has been a real honour.
I am very proud of our data engineering guild and what we are achieving to improve data for our clients, the QuantumBlack culture I have helped to shape, and the opportunities I have helped to create for data engineers to work on interesting and meaningful problems using the latest technology.
Tell us about a career goal or a purpose for your organisation that you are pursuing?
A personal mission of mine is bringing more data science into data engineering to improve and optimise bringing complex data together and developing data pipelines. Cleaning messy data can be frustrating and there are great tools out there that can support. I am keen to augment these with open-source frameworks to help our clients accelerate their own data challenges.
My personal mantra is how do we automate the boring stuff so we can focus on the fun stuff? That is, identify where there could be more value through new and interesting datasets, features, and faster development/deployment through DataOps initiatives.
How closely aligned to the business are data and analytics both within your own organisation and at an industry level? What helps to bring the two closer together?
A big part of our work with clients is helping bring the business closer to data and analytics. In many organisations data is still seen as a back office function, closer to IT, rather than a strategic asset that will drive significant value. Especially in relation to data science solutions, business ownership and involvement is critical to ensure accurate outcomes and adoption – the "last mile" of analytics.
A cross-functional iterative approach from the start, involving business leadership and stakeholders, taking them on the journey helps build trust and understanding. Turning up with a black box analytical model built in isolation only fosters confusion and mistrust.
We often recommend developing or hiring “translators”, specialist roles that can bridge the business and analytics: they understand the business problems, how to translate them into analytical problems (together with data scientists), and how to meaningfully articulate analytics output back to the business and what it means in terms of performance improvement.
Without these bridging roles it can be frustrating for business subject matter experts and technical data engineers/scientists to collaborate effectively and push analytics successfully through to production and actually solve their intended problem (and generate the value).
What is your view on how to develop a data culture in an organisation, building out data literacy and creating a data-first mindset?
Our main success stories have probably had three common components:
I often see organisations that want to wait for the perfect foundations or technology environment to be in place first; actually, you often do not need a huge amount of investment to demonstrate real value from data and analytics. Start small, start building quickly (and safely), and learn through that process.