David Pidsley

Medallion Architecture: Bronze. Silver. Gold. STOP THERE? Think again. In a world of real time decisions, hyper-personalisation and AI driven ops - Gold just isn’t enough. Welcome to the Platinum Layer in Databricks' Medallion Architecture. It’s not just another layer, it’s a leap toward precision, performance and AI readiness. We’re talking: ⚡ Sub second latency 🧠 Agentic AI chatbot integration 📊 Ultra refined, predictive data models 🔐 Centralised governance with Unity Catalog This isn’t a "nice to have." For businesses aiming to lead, the Platinum layer is a strategic necessity. We break it down in our latest Insight: Use cases, when (and when not) to go Platinum and how leading enterprises are doing it. Read the full article: https://lnkd.in/d2BgMXHG

13

Unified AI platforms, graphs, and converged fabrics are powerful. But if you can’t see beyond your direct vendors, you’re building intelligence on top of blind spots. 77% of CISOs rank third-party cyber risk as high impact, yet only 15% have full supply-chain visibility . That’s not innovation friction. That’s GIGO at scale — and the beginning of a FrankenStack. Gareth’s Three Bullets are directionally right. Snowflake is pushing unified AI data platforms. Neo4j attempts grounding AI in a truth layer. Gartner signaling convergence. But here’s the structural tension underneath it all: 𝗚𝗮𝗿𝗯𝗮𝗴𝗲 𝗜𝗻. 𝗚𝗼𝘀𝗽𝗲𝗹 𝗢𝘂𝘁. That’s the 2026 version of GIGO. AI doesn’t fix blind spots. It industrializes them. A unified platform is powerful. But if your “unified” layer only unifies first-party and direct vendor data, you’ve optimized the visible 15% and automated ignorance across the other 85%. That’s not transformation. That’s acceleration layered over opacity. 𝗔𝗜 𝗮𝗺𝗽𝗹𝗶𝗳𝗶𝗲𝘀 𝘄𝗵𝗮𝘁 𝗶𝘁 𝗶𝗻𝗵𝗲𝗿𝗶𝘁𝘀. If it inherits partial visibility, it scales partial truth. The Neo4j case study is compelling because it introduces defensibility. A graph as a “truth layer” is exactly how you anchor AI in relationships instead of fragments. But truth is not just entity linkage. Truth is lineage depth. If nearly half of incidents originate beyond direct third parties , your graph must model the extended enterprise — not just the org chart and primary vendors. Convergence is strategic. But without classification discipline, convergence creates something else: 𝗙𝗿𝗮𝗻𝗸𝗲𝗻𝗦𝘁𝗮𝗰𝗸. A stitched-together ecosystem of: • GRC platforms that weren’t built for continuous third-party monitoring • AI agents layered on incomplete data • Questionnaires that 71% of CISOs say aren’t effective enough • Dashboards that look unified but aren’t structurally integrated It looks modern. It demos well. It audits poorly. Before you converge, you must control. 𝗥𝗮𝘁𝗶𝗼𝗻𝗮𝗹𝗶𝘇𝗲. 𝗖𝗹𝗮𝘀𝘀𝗶𝗳𝘆. 𝗔𝘁𝘁𝗿𝗶𝗯𝘂𝘁𝗲. 𝗩𝗮𝗹𝗶𝗱𝗮𝘁𝗲. Without that sequence, your data fabric is just an expensive blanket thrown over unmanaged risk. The CISO survey highlights something even more uncomfortable: Regulatory pressure is rising for 62% of organizations . Visibility is not. That delta is where enforcement lives. So the real question behind Gareth’s bullets isn’t: “Are you adopting unified AI, graph truth layers, and DMP convergence?” It’s: 𝗔𝗿𝗲 𝘆𝗼𝘂 𝗯𝘂𝗶𝗹𝗱𝗶𝗻𝗴 𝗼𝗻 𝘃𝗮𝗹𝗶𝗱𝗮𝘁𝗲𝗱 𝗴𝗲𝗼𝗺𝗲𝘁𝗿𝘆 — 𝗼𝗿 𝗮 𝗙𝗿𝗮𝗻𝗸𝗲𝗻𝗦𝘁𝗮𝗰𝗸 𝗼𝗳 𝗮𝘀𝘀𝘂𝗺𝗽𝘁𝗶𝗼𝗻𝘀? AI without custody is scaled GIGO. Convergence without classification is consolidated chaos. And blind spots don’t shrink just because you put them in a platform. "Control the inputs. Or prepare to defend the outputs. Geometry first. Everything else second." — Martin #DataArchitecture #AI #GIGO #FrankenStack #ThirdPartyRisk #Governance #ExtendedEnterprise #TTB #3DI #RCAV

2

The UK is sitting on one of the most valuable datasets in the world and most of it is still locked away in government systems. But the National Data Library project could change all of that. Whether it becomes another well meaning initiative or a genuine economic catalyst depends on what happens next. Here’s my take on why this moment matters. #Data #NDL #UKGovernment #Datasets #PublicData #Economy #UK

3

Why do AI models overuse em dashes? I asked Claude, ChatGPT, and Gemini the same question. All three gave surprisingly similar answers that converge on: AI writing optimises for perceived quality over actual quality. Key reasons: > Cognitive efficiency - Em dashes let models add nuance without committing to stricter punctuation rules (semicolons, parentheses, new sentences) > Training data bias - High-quality writing in our training sets (journalism, essays, long-form content) overuses them compared to casual human writing > RLHF pressure - During training, humans rated "sophisticated-sounding" responses higher. Em dashes = sophistication signal, so models learned to use them liberally > Safe hedging - They allow nuance and personality without the risk of jokes falling flat or opinions causing offense > Information density - They're perfect for nesting secondary explanations without breaking sentence flow The irony - All three models used em dashes heavily while explaining why they overuse em dashes.

13

1 Comment

Alan Turing’s Legacy — and the Future of British AI Leadership Alan Turing was a technological genius — a pioneer of artificial intelligence and creator of the “Turing Machine,” the conceptual foundation of modern computing. “The Prof” set a high standard for intellectual excellence from Old Blighty. It is deeply concerning that Britain, once truly Great, now has an institute bearing his name that is failing to deliver the world-class contributions urgently needed in the age of artificial intelligence. (See BBC News article below 👇) As a Brit, I have long honored the genius and memory of this extraordinary man through two internationally recognized #EthicalAI innovations: 1. The Ethical AI Certification and Maturity Model™ — cited by NASA. 2. The New Turing Test for #AGI — demonstrated publicly for the first time with citizen participation and assessment I have tracked The Alan Turing Institute for years. While a few valuable white papers have emerged, the overall contribution to advancing British AI leadership has been, at best, mediocre. It pains me to say this. The Cultural Challenge: The quality of leadership shapes the quality of an organization’s culture — and culture determines the quality of its results. This is business 101. It appears the Institute’s challenges are not only strategic, but cultural. This is fixable — but not by leadership change alone. It requires a complete mindset shift, top to bottom, anchored in a clear vision and purpose that redefines the Institute’s raison d’être. The Broader National Context: The decline in the Institute’s performance reflects a wider malaise. The British government appears to have lost sight of the values and principles on which this nation was built. Policy seems too often driven by ideology — such as the fixation on Net Zero — at the expense of addressing pressing realities, from supporting our elderly through the winter to tackling the cost-of-living crisis. Strong leadership is about courage, clarity, and conviction. Alan Turing embodied these qualities fearlessly — and institutions that carry his name should do the same. The House of Lords remains one of the few bastions of common sense, upholding British values and holding government to account. A Call to Action: Britain has the potential to lead the world in the age of intelligent machines. My hope is that the Alan Turing Institute transforms itself into a true beacon of innovation and leadership — one that honors Turing’s genius, embraces ethical principles, and contributes to global advancement. The world needs Britain to rise again in AI leadership. Will we hear the call? Zoe Kleinman Katalin Bártfai-Walcott Peter R. Lord Holmes #AlanTuring #GreatBritain #AI #Leadership #ThinkingMachines

4

2 Comments

Another nugget from the CGT supply chain technology survey. AI is top of mind, of course - but everyone should be careful to understand what "AI" means in their software and stack.

5

Transforming mission impact through modern data science. The British Heart Foundation teamed up with Simpson Associates to build a scalable, secure Data Science Environment using Databricks on Microsoft Azure. This powerful platform is helping the British Heart Foundation: -           Automate data workflows and reduce manual effort. -           Make faster, evidence-based decisions. -           Enable real-time collaboration across teams. Watch the full recording from our webinar: https://lnkd.in/eCtUgQsr #Databricks #Azure #DigitalTransformation #BritishHeartFoundation

6

1 Comment

Organisations require more accurate answers, greater trust and smarter decisions from their AI models. Graph RAG improves AI reliability by grounding responses in structured, verifiable knowledge — reducing hallucinations and building confidence across your organisation. Watch this video by Sean Martin explaining how this can be achieved. https://lnkd.in/g_fvSSZx #AI #GraphRAG #TrustworthyAI #GenAI #EnterpriseAI

24

🚨 Here's a sobering stat: 60-73% of enterprise data goes unused for analytics & AI.* With Snowflake Cortex AI now GA, the barrier to AI implementation is lower than ever. But there's still one critical challenge: getting your diverse data INTO Snowflake in the first place. Our latest blog explores how CData's Cortex AI integration toolkit is solving this exact problem—with real customer results like: ✅ 90-minute data lag → near-realtime ingestion ✅ 80% pipeline automation ✅ Seven-figure inventory cost savings Get ready to unlock your unused data for AI: https://bit.ly/3I9Us4F #Snowflake #AI #DataIntegration #CortexAI #DataPipelines *Forrester report.

3

This was my favourite quote from the recent interview on the excellent Joe Walker Podcast with Francis Fukuyama on AGI and the Recommencement of History. Francis Fukuyama’s remark made me reflect that the way we frame “intelligence” shapes how we think about AI. If we treat narrow, technical reasoning as the ultimate human virtue, we risk sidelining the other types of intelligence (e.g. emotional, political, cultural) required for individuals, organisations and societies to genuinely thrive. Arguably, these types of intelligence become more important in a world where analytical intelligence is commodified by AI. I’d be interested in hearing your thoughts on the role different types of intelligence play in the development and governance of AI. Christina Wiremu-Brook Joel Backwell Paul Hubbard

21

2 Comments

Great to see James Garner and Yoshi Fabyan Soornack doing a Project Flux live session at the APM data and AI conference. Responsible use of AI. Should this be managed by a central team or federated? Ask a data person and they want to grow an empire. Ask a project person and they want to take control of their own destiny and leverage the latest tools. Shouldn’t ethics and responsible use be embedded skills? Like GDPR? Shouldn’t management set the boundaries and interpretation rather than a separate team - this is a legal and fiduciary responsibility. There is also a tension between going fast and breaking things vs going slow and doing it safely. I sense there is a middle ground. Set a team of trailblazers to go ahead of the pack. Give them safe access to tools and ensure that they are competent. Use them to create a buzz and pipe clean processes. Then put the systems and guardrails in place to formalise. If you try and regulate for everyone it slows everything down. Doesn’t it need layering?

41

4 Comments

As Head of AI at Hutcheon Mearns, I'm thrilled to share how our innovative AI solutions are making a significant impact for our clients. Our initial use cases have demonstrated that effectively implemented AI can be pivotal in enhancing business improvements, streamlining operations, and surpassing competition. By automating repetitive and time-consuming tasks, our AI services free employees to focus on high-value activities that truly contribute to a company’s growth and success. This shift not only boosts productivity but also empowers teams to engage in strategic customer interactions and explore new revenue-generating opportunities. The early results from our clients show that those who embrace AI are not just keeping pace with their competition; they are exceeding it. Companies that leverage our technology are gaining a competitive edge, unlocking efficiencies, and rapidly driving value-add activities that matter most. I invite businesses to connect with us to discover how our tailored AI solutions can enhance your operations, drive efficiency, and put your employees back at the forefront of business-development initiatives. Together, let’s harness the power of AI to propel your organization to new heights! #AI #BusinessImprovements #Efficiency #ValueCreation #HutcheonMearns #CompetitiveAdvantage #KnowledgeManagement

49

Great article from Forbes and Dr Sanjay Kumar. "data is the basis that gives AI its intelligence, rather than an afterthought" https://lnkd.in/eP-wyxvy

7

4 Comments

Sankar makes an excellent point about agents allowing foundational redesign of processes. How will this alter operations processes and supply chain configurations? #agentic #designthinking #Supplychain #Configuration Production and Operations Management Society (POMS)?

13

3 Comments

The UK’s AI for Science Strategy is a big step forward and rightly focuses on research, data, technology and infrastructure. But after spending the last 8 years working in this space, I’ve seen first-hand that real success doesn’t come from models or compute alone, it comes from application, adoption, and most importantly, the people who make that happen. Pillar 3: People and Culture might actually be the most overlooked part of the strategy, and the most important. In my experience the most successful projects have  - Builders and researchers developing the fundamentals - Translators who bridge AI approaches with real industrial problems - Adopters / champions who roll up their sleeves and deploy this technology on the ground Without all three, even the best-funded AI will stall in the 'science'. Collaboration and a genuine growth mindset across industry will determine whether this strategy creates papers or progress. https://lnkd.in/eqCM5_Q2

33

4 Comments

Working in local or national government? Here are 10 reasons why Cadence is the best data platform you’ll ever use: 1) Over 3,000 UK datasets pre-loaded. Extensive data from ONS, OS and government departments mapped and at your fingertips 2) Best-in-class visuals and speed. Low latency, high performance loading of all the data you need. 3) Easy to use. Learn everything you need to know in 5 minutes. So easy my 10-year old can use it. 4) Charts and maps. Easily add charts to maps with a single click. Save hours sourcing, wrangling, processing and analysing data - everything you need is already built in. 5) Planning Ready. OS Basemaps and colour schemes to create MHCLG-compliant planning maps. 6) Analysis made easy. Create core strategy data in minutes. Templates enable creation of strategic evidence for projects like LCWIPs, LAEPs and many more. 7) Create and share geospatial data. In built tools to create map-based data as easily as drawing on a map. 8) Designed for stakeholders. Multiple ways to share with stakeholders including dashboards and stories. 9) Advanced tools designed for your teams. Tools such as isochrones and population analysis made easy. 10) Secure and cloud-based. No infrastructure needed and web-based collaboration. Data stored in UK/EU and fully secure. And a couple more for luck: 11) Text based search. Easily locate the data you need with text-based data search and multiple ways to filter / explore. 12) Money saving guarantee. We’ll work with you to optimise your licenses so that integration of Cadence also delivers overall cost savings. We’re so confident we can save you money we’ll give you a license free for a year if we can’t. If you want to learn more, reach out to me today or try for free at the link below: Thomas Baird Suzanne Conn Bethany Taylor Tony Harbron Holly Whitehead Local Government Association Department for Transport (DfT), United Kingdom

19

1 Comment

Ever since Sadie St Lawrence and I connected years ago for some collaboration with Women In Data™️, I've always admired her passion and expertise for data science and AI, and it's been really inspiring to follow her journey. This is why I was so excited to see her publish the book "Becoming an AI Orchestrator" and as expected it was a great and insightful read for many reasons. As a fellow musician, I really enjoyed the music analogies of how knowledge workers need to shift from being a musician to be an orchestrator, and why being an "orchestrator" is more critical than being a "conductor" in todays times. Other highlights for me were: 👉 The overview of Gen AI application categories across creative tasks, analytical tasks and hybrid tasks 👉 The implications of different types of intelligence for our relationship and interactions with AI 👉 The concrete examples of applying AI in different business functions 👉 An iterative process for AI adoption that starts with identifying human tasks 👉 How to get started in defining the right AI guidelines and policies, including an important part about how to address Bias 👉 A playbook on how to thrive in the age of AI by following certain golden principles Across the entire book Sadie mentions lots of her experiences and stories of how she uses AI in her day to day, which together with other parts shows us a truth that I couldn't resonate more with: doing the right and valuable things with AI starts with every single individual. And reading this book will definitely help you do that. I highly recommend this book to anyone who wants to know how to use AI to enrich their life, their career and our collective future - which ideally should be everyone. #aistrategy #book #tiankaistuff

87

4 Comments