Conrad Annan
I spent 10 years at a FTSE-listed retailer learning what makes AI actually work. Now I help others do the same.
Most AI projects fail because nobody measures whether they work. I fix that.
I help organisations adopt AI with confidence — moving from pilot to production with measurement frameworks that prove value. My approach is built on 15+ years of product leadership at FTSE companies, 50+ automations delivered, and a proven methodology for measuring real outcomes.
At a FTSE-listed retailer, I lead AI product strategy for one of the UK's largest online platforms. Outside of that, I work with organisations who want to use AI strategically — not just adopt tools, but build the capability to measure whether those tools actually work.
Track Record
- Senior Product Manager at a FTSE-listed, £3.5B e-commerce platform serving 26M+ monthly active users
- 10 years at a single company: Agile Delivery Manager → Platform Lead → Senior PM → AI Product Strategy
- 86 epics delivered across Order Management, E-Commerce Foundations, and Digital Core Services
- Directed 5 engineering squads (20+ engineers), owned £5-10M annual technology budget
- Delivered £2.8M measurable impact: +£1.5M conversion, £1.3M cost savings (Ratings & Reviews, Discovery to Live)
- Shipped Premier Delivery Savings, Partner Fulfils (Inditex), Guest Checkout, Limited Drops, Buy the Look (+£2M revenue)
- Presented Returns strategy to EVPs of Product & Technology — programme-level ownership across the organisation
- Architecting AI-generated product content across 41,000+ products — targeting £5-10M annual return rate savings
- Led mobile-first transformation — iOS app releases (RT24-28), Android parity, native returns experience
- Led platform migration from monolith to microservices; CI/CD from 2-week cycles to daily releases
- Built and operate a 350+ AI agent system in daily production use — not theoretical, lived-in daily
What I Believe
Problems first
I learned this the hard way after watching a £200K platform nobody used at my first job. Most AI initiatives fail because they start with technology. I start with what's broken.
Measurement
If you can't measure whether AI is working, you don't know if it's working. Every engagement includes A/B testing, statistical validation, clear success criteria.
Capability
I build your team's ability to continue AI work after I leave. Knowledge transfer is not an add-on — it's embedded in every engagement.
How I Work
These aren't theoretical principles. They're patterns I've validated through enterprise-scale AI deployments. IDC research across 4,000 organisations confirms: firms that measure first achieve 2.8x ROI. That's the approach I bring to every engagement.
Sequential agents beat orchestration at scale
Orchestrator patterns hallucinate when they chain too many decisions. I use sequential agents with structured handoff — each agent has one job, passes a file to the next. Validated across 27+ enterprise workshops.
Every agent needs three things: identity, goal, process
Without a clear persona, a defined objective, and an explicit work-through process, AI agents produce generic output. The specificity of the brief determines the quality of the result.
Constraints protect the work, guardrails protect the system
Constraints define what the agent can and can't do. Guardrails protect against context overflow, hallucination, and drift. Most AI failures come from conflating the two.
Language precision changes everything
The difference between “do not” and “you are forbidden to” changes model behaviour measurably. Careful verb selection, explicit formatting instructions, and structured context architecture are what separate tools that work from tools that almost work.
15+
Years Product Leadership
50+
Automations Delivered
£2.8M
Revenue Impact Delivered
FTSE
Enterprise Experience
Building AI Capability, Not Dependency
Every engagement is designed to leave your organisation stronger — with skills, not just tools.
Knowledge Transfer
Every engagement includes structured capability uplift. Your team leaves able to continue AI work independently.
AI Skills for All
AI upskilling programme already deployed to product teams. Repackaged for organisations at every stage of their AI journey.
Local Founder
Based in High Barnet, London. Father of four. The outcomes of AI transformation directly affect my family and community.
Measurable Impact
Every social value commitment comes with a measurement plan. Impact is tracked, not promised.
Beyond Work
Work is one dimension. Here are the others.