
Make the page
come alive.
My children love their Yoto players. I’m making stories that bring new vocabulary, spelling and school topics into the worlds they listen to.
AI at Benefit Street Partners · Personal projects
This illustrated sequence connects six projects. Storytelling puts vocabulary into context. Clock learning lets children practise moving the hands. Lyonheart is a guided-intake prototype supporting a human conversation. Value Insights connects business information. A fictional company diagram starts with North Star Group, Harbour Studio at 100% ownership and Fieldwork Labs at 75%, then expands to 95 companies. A generated road scene distinguishes a pothole, a puddle and a raised bump. Use previous and next to explore at your own pace, or pause to stop movement. The illustrations are silent.

My children love their Yoto players. I’m making stories that bring new vocabulary, spelling and school topics into the worlds they listen to.

A small game for practising how to tell the time. Move the hands, check an answer and try again.
Try the clockA prototype for gathering information before an assessment. People answer at their own pace; a human practitioner reviews what they share.
At PwC, I helped build a platform that connects financial and operational data so businesses can understand what is driving performance.
Microsoft’s customer storyTurn ownership information into a clear company diagram that people can discuss, edit and use in a presentation.
Build a diagramI’m exploring how dashcam footage could become useful road reports, starting with photo assessment and drafts for a person to review.
Explore the example
Road reporting concept: capture an image, distinguish the feature, review the evidence and prepare a report. This interactive example uses a generated photograph and sends nothing.

Benefit Street Partners
I work on practical uses of AI for investment teams.
Nine years spanning assurance, deals analytics and AI. I helped design, develop and bring Value Insights to market.
Microsoft’s customer storyMy PhD at Reading used machine learning and optimisation to work out complex crystal structures from X-ray measurements. That combination of scientific modelling and testing against evidence still shapes how I approach financial analytics and practical AI.
My research used genetic algorithms, applied machine learning, global optimisation and dispersion-corrected density functional theory (DFT-D) to investigate and solve complex crystal structures from powder X-ray diffraction (PXRD) data.
I teach practical AI and analytics for finance and entrepreneurship. Outside work, I make stories for children and tools for learning and everyday problems.
A proposal for giving vendors realistic, flawed data shortly before a demo, to see what their systems can actually deliver.
LINKEDIN POSTWhat happened when I asked ChatGPT and Claude to build an Excel tax-planning model, including the parts that still needed checking.