My Experience
European Digital Agency
Business Transformation Programme
Context: My client wanted to make systematic use of their data to find clients, organise information, create high-quality written and image content using generativeAI, and automate routine workflows. They also wanted to introduce Agile working to execute projects a more responsive manner.
​​​​
Scoping: I worked with C-level decision makers to understand their problem, investigate potential technical and non-technical solutions and deliver a project proposal that involved building a customised business process automation (BPA) suite using low-code/no-code tools (Bubble, Zapier), to re-engineer several key workflows. The senior leadership approved this plan and asked me to deliver it.
​​​
Delivery: I worked with the internal technology function to design the BPA platform, then project-managed the delivery of this software. This involved eliminating dependence on desktop tools (e.g. Excel, Dynamics). I also acted as subject matter expert for the introduction of LLMs and ChatGPT into the business, and the platform enabled automated content generation by integrating the OpenAI API. I also introduced Agile working practices into the business and modelled effective ways of using these to improve delivery.
​
​​​​
Impact: The project delivered a software suite that enabled the firm to search, share, understand and catalogue firm-wide data in a single location, automate a large range of sales and marketing tasks and improve data quality and availability. As it heavily used no-code tools, the software was easily maintainable by the technical team. The firm estimated this saved thousands of hours of wasted effort annually. Agile was also well embedded within the business, with several teams actively using it for project delivery.
UK Hedge Fund
Data & AI Platform Engineering
Context: My client was a trading desk spinning out of an established hedge fund. They wanted to replicate their institutional quality trading and analytical infrastructure, as well as improve on it using market-leading commercial software, and automate large parts of their own investment process that currently involved manual execution.
​​
Scoping: I worked with the portfolio manager and senior traders to understand their requirements in detail, assessed the technical challenges, and proposed a platform architecture to meet their needs. This composed a trading system, research infrastructure, risk dashboard, and an integration platform to handle data from dozens of diverse sources, including fundamental, macro-economic and market data.
​
Delivery: I project-managed the delivery of the system while also acting as technical lead. I built a small team of senior software engineers to assist me in delivering the platform. The system enabled analysts to access and query large quantities of historic data with minimal technical knowledge, and used automated machine learning to help analysts find patterns in the data. It also incorporated a sophisticated risk dashboard and low-contact trading using a system of price alerts on pre-specified price or indicator levels, meaning trades could be executed rapidly when market conditions aligned.
​​
Impact: The team were able to move almost seamlessly onto the new system ahead of schedule, with minimal downtime or disruption. The system made use of open source components and public frameworks, to minimise vendor lock-in, ensure it was easily maintained and the functionality was extensible. By the close of my engagement, the team was planning a custom mobile app to complement the platform and enable prices and positions to be checked, and trades to be made from a phone.
​I've held technical and leadership positions in industries from energy to finance to the public sector, and worked in roles from data architecture, to machine learning engineering, to technical programme management. I have a proven track record leading and scaling teams, managing complex and multifaceted technology programmes, and successfully engaging C-level stakeholders.
This is a short selection of my previous work.
Global Energy Provider
Innovation Programme
Context: The company was seeking to use data more strategically to improve its operations. I was engaged as the technical director and lead architect of a major programme of work that involved building a cross-functional data & analytics Centre of Excellence from scratch to over 100 FTE.
​
Scoping: After consulting widely with senior stakeholders to understand their key commercial priorities, I led a rapid programme of requirements gathering to identify key user needs through work shadowing and interviews. After identifying some high-impact use cases, I rapidly built an analytical dashboard which clearly showed the potential for seven-figure annual cost reductions based on a conservative analysis. Individual projects would include better use of customer data to reduce customer debt, and reducing customer bills by using smart meter data. Based on this, I built a business case which scoped potential solutions and won approval for a major multi-year buildout of the the firm's data science capability and the Data & AI Centre of Excellence within the firm's Innovation Hub.
​​
​​​​​​
Delivery: As lead architect. I designed and supervised the build of an AWS data lake for large-scale (multi-PB) data stores, for data from customer interactions (phone calls, web chat, email), as well as geographic (GIS) and open source data. I also advised project leads on elements such as appropriate database schemas and aspects of the machine learning models that underpinned several projects. I further led the recruiting and professional development process for technical staff, including data engineers and data scientists, conducting interviews and supervising and mentoring junior team members.
​​​
​
​
​
Impact: The Centre reached full operating capability in under 2 years, ahead of schedule. Several individual projects had seven-figure P&L impact to the firm's bottom line, and the Centre was moved from the Innovation Hub into Technology division, to catalyse change, diffuse new skills and drive Agile working throughout the organisation.
UK Government Department
Digital Transformation Programme
Context: I was technical delivery lead for the data & AI workstream in a major digital transformation programme undertaken by a large UK government department. This was a high-visibility role as better use of data was central to achieving the ambitious spending reductions the department had committed to.
​​​
Scoping: The workstream was composed of several major projects, including Single Customer View (SCV), which was intended to give the department an accessible, up to date and all-round view of each of its users for the first time, breaking down prevailing data silos. A second project involved re-engineering the analytical counter-fraud hub to move from a SQL-based business rules-driven process to a machine-learning focused approach. A third was to provide supporting analysis for a programme to downsize the department's office footprint, while minimising the impact to millions of service users and and 80,000 staff. I also had responsibility for leading the development of Natural Language Processing (NLP) capability in government. This was a key priority due to the very large volume of written material produced by public services.
​​​​​
Delivery: I wrote the technical specification for the SCV model, involving integration of dozens of variable-quality data sources using fuzzy matching techniques. I also developed an analytical framework that supported the departmental facilities programme by using Google TravelTime API and GIS data alongside a custom optimisation engine. I also prototyped a novel analytical approach for the counter-fraud hub that involved using neural networks to predict fraud using sequences of interactions with the department. Simultaneously, I designed the architecture for a new cloud data lake, and was heavily involved in recruiting and talent development to grow the team and share knowledge . ​​​​
​​
​
Impact: The Digital Innovation Hub was scaled to over 50 technical staff in under 15 months, contributing novel and cutting edge analyses that were used almost every day by every departmental employee, and providing critical decision support for several billions of pounds of annual spending. The SCV alone was estimated to have saved many thousands of hours of data integration in a single year. The departmental facilities footprint was reduced by almost 20%, with minimal impact on service users and staff