Hodman Murad addresses the "scale paradox" of modern data management: the more data you have, the harder it is to find what matters. Traditional catalogs fail because they rely on automated crawling, which creates a noise-to-signal ratio that practitioners eventually ignore. Murad argues for a paradigm shift, implementing a three-part framework to ensure the catalog remains a trusted, load-bearing system.

In this post, Charlie recalls watching a team get defunded despite building a recommendation engine that could’ve saved $2M annually. He stresses making data products visible, accessible, and valued by the business.

This episode explores how enterprises are actually implementing AI beyond the hype. He discusses the importance of strong data foundations, targeted AI use cases, business alignment, leadership, and change management for scaling AI successfully in large global organisations.

This executive AI roundtable with Microsoft, C3 AI, and Fractal on scaling Enterprise AI shares insights on how companies can move beyond AI hype to measurable business value, while avoiding large “big-bang” AI programs and instead scaling through small, fast wins. It also had lessons on building data foundations and AI capabilities simultaneously, and enabling the turning of AI experimentation into repeatable, scalable enterprise outcomes.

Piotr Jurowiec argues that the software industry is transitioning from the "Copilot Era" of AI assistance to an Agentic Era, where AI doesn't just suggest code but autonomously plans, executes, and heals systems using OODA loops. This shift fundamentally alters the physics of the SDLC.

Cassie Kozyrkov explores the jarring "trust gap" in modern enterprise AI, where 86% of decision-makers plan to increase agentic exposure despite only 6% currently trusting agents with end-to-end processes. She argues that the primary bottleneck for AI success has shifted from the models themselves to the "plumbing" and the people.

In this LConnect podcast, Marc reflects on his cross-industry supply chain journey and shares a key leadership lesson: “go slow to go fast,” emphasizing thoughtful decision-making over reactive speed. He also offers a pragmatic view on AI and logistics, highlighting that real impact comes from using data and automation to improve decisions, not chasing hype.

Irshad Saifi argues that the future of leadership is not a battle between humans and machines, but a synergistic fusion of Human Intelligence (HI) and Artificial Intelligence (AI). While AI excels at speed, scale, and precision, Saifi identifies three core human "multipliers" viz Empathy, Curiosity, and Humor that remain irreplaceable.

Jessica Talisman argues that the systematic capture and explicit representation of process knowledge is a critical necessity for enabling trustworthy and auditable AI systems. While traditional process knowledge often remains tacit the solution lies in integrating process engineering principles with semantic technologies to encode rich context and meaning.

Data stewardship is likened to the noble care of Gondor’s realm, arguing true stewards watch over data as a trusted asset for the whole organization, not just individual job tasks, inspired by Tolkien’s metaphor.

Ali Azhar argues that synthetic data is filling the gap left by the constraints of real-world data, which lacks the volume and quality necessary for modern AI to cover every rare edge case and scenario. Synthetic data allows developers to create virtual worlds and failures on demand, shifting data from a constraint to a creative tool that enables faster iteration and model performance improvement.

Rob Manson argues that the seemingly inconsistent behavior of LLMs stems from arbitration, the token-by-token, internal process where the model implicitly decides how much to trust its three competing forces: latent states, routes, and anchors. This arbitration is an emergent property of how attention heads and MLPs interact within the residual stream scratchpad, a high-dimensional vector space.

William Inmon argues that data models act as essential "lighthouses" for navigating the chaos of corporate data, distinguishing between inward-facing structured models and outward-facing textual models. He emphasises that while structured models focus on stable keys and attributes, textual models must uniquely account for context.

Dia A. argues that competitive advantage is shifting to organisations that move beyond forcing data leaders to justify their value and instead fully integrate & empower them at the heart of their strategy. Historically, data leaders were defensive, but the most successful companies now embed them across functions to accelerate decision-making, break down silos, & drive quantifiable outcomes.

In this fireside chat, François Zimmerman and Sergej Barkar explored how Tableau and Google Cloud work together to scale modern analytics. Sergej shared a real-world shift from centralized BI to governed self-service, highlighting cloud-native deployment, scalability, and democratized analytics.

Anna Bergevin argues that data platform work remains invisible because platform teams, unlike SaaS product teams, lack product marketers. This results in low adoption, redundant work, waste, and strategic vulnerability during budget cuts. The PM must own this internal GTM gap, viewing evangelism & storytelling as an essential part of shipping.

Bassel AbdulHak explains that Slowly Changing Dimensions (SCDs) are a critical data warehousing concept used to manage & preserve the history of dimension data that changes over time, albeit slowly. He details five common types, including Type 2 and Type 4.

This article talks about how AI-driven BI enables personalised answers, but risks multiple versions of truth. The real test shifts to semantic layers, including metadata, definitions, and guardrails, to balance flexibility with consistency and maintain trust.

This in-person event explores AI risks, opportunities, and how authenticity, relationships, coaching & networking help us thrive in an AI-augmented world.

Darragh Murray explains that the semantic layer formalises institutional business knowledge, providing clear, shared definitions of data fields, metrics, and labels that typically reside only in the heads of senior analysts. This layer is now essential AI infrastructure because generative AI (LLMs) requires this structured business context to generate accurate insights.

In this post, Charlie argues that many companies call themselves data-driven but aren’t, and dashboards alone don’t equal impact. So true data culture means decisions start with data, not just justify choices made.

Vlad Radziuk explores the limitations of OWL-based ontologies, arguing they lack the expressiveness to model real-world complexity, such as a holon or hypergraphs. This limitation stems from OWL being based on First-Order Logic, which enforces structural constraints like prohibiting "type-instance collapse," requiring awkward workarounds like reification.

The article explains how data domains create shared business meaning, ownership, and governance across enterprise data ecosystems, helping organizations improve analytics, AI readiness, data quality, and cross-functional alignment.

In the Forbes interview, Maria Demaree discusses leading large-scale digital transformation while fostering a culture of purpose and mission focus, balancing operational excellence with strategic modernisation.