Big Data LDN 2025: Key Talks, Top Takeaways & Next Steps for Data & AI Leaders

Big Data LDN celebrated its 10th edition on 24–25 September 2025 at Olympia London, attracting thousands of data and AI professionals. Across 300+ speakers and 16 theatres, the show spotlighted how organisations are moving from analytics to intelligent action.

The headline topics were:

• Agentic AI and autonomous data workflows
• Unified data platforms and data products
• Governance-by-design and ethical AI
• Real-time decisioning, observability and AI safety

“AI-powered is now everywhere… this year, ‘agentic’ had obviously joined most offerings.”

Keynotes that Set the Agenda

Google Cloud: The Agentic Era

Google Cloud explained how data science is shifting to autonomous agents that reason and act across enterprises. Their call to action: design for unified, AI-native platforms because silos can’t sustain intelligent agents.

What you can do: Identify five workflows where agents could boost productivity (e.g., pipeline remediation, cost governance) and create an agent safety spec with approvals and audits.

Professor Brian Cox: The Universe as a Quantum Computer

Professor Brian Cox reminded the audience that complexity, measurement, and physical limits shape how we compute at scale - a timely warning as AI ambitions grow.

Women in Data: Ethics at the Core

Sessions from Women in Data highlighted bias, harm modelling, and real-world impact. For many, this signalled a shift: ethics is now a core product requirement, not a compliance afterthought.

“This year’s talk… pulls together purpose, adaptive mindsets, and effective governance… speed without proper foundations is a liability.”  Jovita Tam



Key Insights from Jez Clark: Culture & Talent Strategy for Business Impact

Jez Clark’s session on Culture and Talent Strategy underscored that people and mindset remain the foundation for every AI and data initiative. Drawing on his long experience at the heart of the data industry, Jez highlighted three key messages:

1. Culture is the hidden architecture of innovation
   Without a shared purpose and psychological safety, even the best technology cannot deliver sustainable impact.
2. Talent strategy drives measurable business outcomes
   Jez encouraged leaders to align data team structures and growth pathways directly with business KPIs, ensuring investment in skills maps to value creation.
3. Leaders must enable continuous learning and adaptive teams
   Building learning loops - regular review, reflection, and re-skilling - keeps teams resilient in fast-changing AI environments.

“People create the conditions for technology to succeed. Culture and talent are not side projects - they are the strategy.” Jez Clark

For a deeper dive, you can read Jez's full presentation here.

Six Takeaways for Data & AI Professionals

1. Build for Agentic AI
   Unify operational and analytical data, embed memory and tool-use interfaces, and implement observability and rollback from day one.
2. Governance by Design
   Embed policies into data contracts and CI/CD, automate lineage-based controls, and generate continuous compliance evidence.
3. Trust as a Product Requirement
   Model harm to users and non-users, publish transparent impact assessments, and keep ethics at the heart of every AI initiative.
4. Ruthless Data Product Management
   Track business value, deprecate underperforming products, and encourage reuse via a semantic data marketplace.
5. Next-Level Observability
   Monitor prompts, tool calls, cost, and guardrail events for both data pipelines and AI agents.
6. Real-Time Decision Automation
   Combine streaming data, rules, and ML for closed-loop operations, with “pause/confirm” safeguards before full autonomy.

Quick 90‑Day Action Plan

• Days 0–30: Create an Agent Safety Spec, audit your top 10 data products, and integrate PII checks into CI/CD.
  
• Days 31–60: Launch two pilot agents (e.g., pipeline remediation and BI Q&A) and implement full observability.
  
• Days 61–90: Promote a successful pilot to production, roll out a data product marketplace, and publish an AI Transparency Note.

Why It Matters

Big Data LDN 2025 made one message crystal clear:  Unify before you amplify.

AI magnifies both value and risk. For data and AI leaders, now is the moment to put strong foundations in place, experiment safely, and scale only what proves real value.

See you there next year!