I spent over a decade helping global brands better understand their audiences, leading major accounts in top creative agencies and working alongside the founders of two insight consultancies. I saw over and over again that the richest insights – the ones that get to the human “why” behind behaviour and decisions – are hidden in messy text data.
Surveys were dumbed down to make analysis easier. Interview transcripts were scanned for soundbites instead of statistical rigour. And when AI came along, the hype was quickly replaced with disappointment. It hallucinates, miscounts, and misses the nuance entirely.
I experienced the same frustration first-hand when I founded my first venture. I taught myself to code on maternity leave, built milk-facts.co.uk to bring better data to a confused debate, and saw it go viral overnight. That grew into Wish I’d Known, a huge crowdsourced database of parenting experiences. It taught me two things:
1️⃣ The power of tapping into collective human experience.
2️⃣ The need for tools that can analyse it accurately and at scale.
Now, as cofounder & CEO of WholeSum, I’m building exactly that. Our platform combines AI, symbolic reasoning, and statistical models to bring the rigour of quantitative analysis to qualitative text. Removing common hallucinations, with a full audit trail, and built for scale. We call it an intelligence layer for ‘human’ data.