Writing
Essays on data, AI, and the people behind both .
Working notes and longer-form thinking from the field. Organised by pillar so you can read by what you actually care about. Published first here, syndicated to LinkedIn later.
Agentic AI
Agentic AI & Enterprise Architecture
Tiered maturity, deployment patterns, the agent stack for organisations that mean it.
- From Assistant to Autonomous Partner: My First Contact With Google's Antigravity Antigravity moves AI coding from autocomplete-on-steroids to agent-native. Notes from a day inside the new agentic development surface.
- My Reflections on the AI Agent Playbook The most successful AI initiatives are not radical overhauls. They are phased: match agents to existing workflows, then transform. Build the central nervous system: observability, accountability, governance.
- From Input-Output to Intent: Architecting the Future of Enterprise AI A three-tier maturity framework for deploying agentic AI in the enterprise: Foundation, Workflows, Constrained Autonomy. Trust before capability.
Governance
AI Governance & EU AI Act
Explainability, accountability, audit trails, regulatory commentary at exec scale.
- Data Governance: The Bedrock of Trust and Value Good governance is not bureaucracy. It is the mint that produces the trust your data-driven mentality runs on. A journey, never a destination.
- Every Day Is Data Privacy Day Data privacy is not an annual celebration. It is a daily operating principle, individual and organisational. A short note from a regulated operator.
- Zero Trust by Design Trust in AI is built when real-world outcomes match predictions. Borrowing the Zero Trust posture from data security, applied to the AI lifecycle.
Infrastructure
Data Infrastructure & Engineering Practice
Real-time inference, multi-cloud, observability, the technical backbone of AI work.
- Serving AI in Real-Time: Architecture for Low-Latency, High-Availability Inference End-to-end serving architecture for real-time AI inference. Managed vs self-hosted, gRPC vs REST, edge vs centralised, the trade-offs that actually matter.
- CTO and CDO Infrastructure Challenges Modular infrastructure, resilient pipelines, future-proofing. The architectural moves that handle AI workloads without overwhelming the budget or the team.
- Sampling Bias is a Good Thing: TikTok's Imbalanced Training Approach Field note on a TikTok and University of Connecticut paper. Non-uniform negative sampling that keeps the confusing samples and discards the easy negatives. Why it matters for any recommendation system.
Leadership
Leadership for CTOs and CDOs
Exec-level guidance for data and AI leaders. How to think about this, not just what to do.
- The Three Legs Every CTO and CDO Stands On When Implementing Data and AI Data, infrastructure, and people. The three legs that carry a real data-driven push at large-organisation scale. Build them together, not in sequence.
- Ten Questions on AI Implementation From identifying the right problem to measuring impact, scaling beyond MVP, and the ethical guardrails: a Q&A on how AI implementation actually plays out at scale.
- Join the Journey: Why I'm Writing This Series Series intro. The buzz says data is the new oil and AI is revolutionising everything. 87% of data projects never reach production. Here is the gap, and what I plan to write about.
- Implementing AI Solutions: The Practical Playbook for Innovators 87% of data projects never reach production. 27% are profitable. Here is what separates the survivors: clarity of purpose, agile process, and people who actually carry it.
Human Factor
The Human Factor
Teams, talent, transformation, capability transfer. Building people before building code.
- The Human Factor in Data and AI Teams, talent, transformation: the three human challenges that decide whether your data and AI programme delivers, or quietly stalls.
- Tailored Communication for Tech Leaders How tech leaders make their message land with non-technical audiences without diluting the vision: ethical storytelling woven with data, and answering 'why should I care' first.
- Beyond Monitoring: Continuous Improvement in Data Governance Training Monitoring is not the goal. Building a continuous improvement engine that turns trainees into data governance champions is. Five principles that move training from event to ecosystem.
- Unleashing the Power of Data: A Mutant's Journey An unconventional reflection on building a data team, told through the lens of Professor X and the X-Men. The team is the mutation; data is the power; democratisation is the legacy.
Field Notes
Field Notes
Conference recaps, tool tests, sector reflections. Lighter cadence, real-time signal.
- Field Notes: London Tech Week and VivaTech 2025 Europe's two biggest tech events sent the same message: AI is no longer a tool, it is foundational infrastructure. CEOs and CTOs, this is the lead-from-the-top moment.
- May the Fourth: What Star Wars Teaches Us About Data Projects MVP focus, governance from day one, and data democratisation, told through the metaphor of the saga's most expansive universe.
- Personal Brand and the Governance Mandate As an exec in data and tech, the brand you build for yourself has to sync with the company's data strategy, not clash with it. A short reflection.
- Crownpeak Customer Conference: Notes From the Panel Short Q&A from a panel on merchant priorities, margin pressure, ChatGPT's mainstream moment, and what to focus on before peak season. June 2023.