Microsoft MVP | AI & Fabric Specialist for Business Transformation | Helping Enterprises Unlock ROI with Azure, Copilot & Power BI | Forbes Tech | Keynote Speaker | CEO, Onyx Data

Strategic Data & AI Leader | Microsoft MVP | Oxford AI Strategist 

I bridge the gap between technical complexity and Board-level strategy. With nearly 15 years of experience, I have delivered over £93M in revenue growth for global enterprises by transforming data estates into high-velocity strategic assets.

Global Authority & Scalable Insight: 

LinkedIn Learning Instructor: Author of "How to Build and Execute a Successful Data Strategy," an industry-standard framework used by 10,000+ global learners (4.5/5 rating). 
Industry Recognition: Microsoft Data Platform MVP, Forbes Technology Council member, and Gartner Community Ambassador. 
Academic Rigor: Currently calibrating enterprise AI risk and leadership at the University of Oxford (Saïd Business School).

The "Onyx Methodology" in Action:  I help organisations move past "AI Hype" to achieve Operational Utility. Whether leading a global practice of 60+ consultants or advising the C-suite on £200TB+ migrations, my focus is on de-risking technology investments to drive EBITDA.

Core Specialisms:  AI Strategy & Governance (Purview) | Microsoft Fabric & Databricks Architecture | Data Monetisation | High-Performance Team Scaling.

Let’s connect,  I’m focused on the intersection of AI Ethics, Enterprise Strategy, and high-stakes Data Leadership. Always open to discussing the future of our industry with fellow Partners and Architects.

Leon Gordon

Think of your Fabric Capacity like a Train. 

Licensing is the #1 confusion point I see with Microsoft Fabric.

You bought a capacity, but your team still gets "Upgrade to Pro" prompts. 

Why?

It helps to visualise it like a Commuter Rail System:

🎫 The Small Train (F2 - F32): Everyone needs an individual ticket (Pro License) to board and look out the window (view reports). It doesn't matter if you own the train; the passengers still need to pay.

🚄 The High-Speed Rail (F64+): This train is so large that the "Venue" subsidises the cost. Standard Pass holders (Free Users) can ride for free just to look out the windows. This is the "F64 Magic Number."

👮 The Conductor Rule: Regardless of the train size, you always need a special license (Pro) to actually drive the train or manage the passengers (share content/workspaces).

I created this cheat sheet to break down the 3 building blocks of Fabric licensing:

1. Capacity (The Engine): Your computing power.
2. Per-User (The Driver): The Pro/PPU seats.
3. The F64 Rule: When free users can finally see content.

Save this image. It will save you from overpaying for Pro licenses you might not need.

#MicrosoftFabric #PowerBI #DataStrategy #DataArchitecture #Azure

Microsoft MVP | Oxford AI Strategist | Interim Technical CDO / Head of AI & Data | Architecting Sovereign Governance & £93M+ ROI for Global Enterprise

Brian Julius

Over 40 years of experience in a wide range of data analytics/data science roles, both as a practitioner and as a senior manager of analytics teams.

I hold a B.A. in Economics and a M.S. in Public Policy and Management, both heavily focused on econometrics and other quantitative methods.  After beginning my career in economic consulting, I worked 30+ years for the U.S. Federal government (NOAA), in a variety of data-driven roles, including: Economist/Econometrician, Business Analyst, Data Analyst/Data Scientist, Testifying Expert Witness, Deputy Chief Financial Officer, IT System Owner, and Project Manager, as well as in a variety of executive/senior management positions.   I also served as PM on NOAA's first large-scale implementation of Power BI and Dynamics.

After retiring from the government in 2021, I served as the Enterprise DNA Chief Content Officer -  oversee all the creation of all free and member-only technical content, and served as an expert and content creator myself.

For the past three years, I have focused my research, consulting, and writing on the intersection between AI and data analysis.  While not an expert in the field of AI, I do consider myself an expert *USER* of of AI in the areas of Data Analysis and Data Science, for example having developed the first AI iterative DAX and M code solvers and have helped numerous organizations and thousands of individuals learn to better leverage AI technology.

I believe that in the AI era, all analysts need to see themselves as managers developing the proper skills to effectively lead a diverse team of models, including what tasks to delegate to AI, and the skillsets to focus on, such that AI becomes a force multiplier, and not a competitor.

I remain a voracious learner and still find there's nothing that compares to the thrill of data revealing something unexpected or counterintuitive about the way the world works.  

Currently, my principal objective is helping people in the data field achieve their own goals.  I strive daily to be a positive contributor to this great community that has generously given so much to me over the years.

Anthropic's Claude Code owns 54% of the AI coding market. But its dominance isn't about being a "better product."  It's about a dynamic between producer and consumer that we've never seen before...

I am a huge fan of Claude Code (CC) for coding, data analysis, automation, and more.  So, it's great to log on every day, see more and more people falling in love with it, using it in incredible ways, and sharing what they learn and produce with the broader community

But beyond CC as just an exceptional tool, there is a much larger dynamic at play - one that is not only redefining software engineering processes, but has reshaped the entire producer-consumer relationship in unprecedented ways

For the first time in history, Producer and Consumer are running on the exact same tool, each leveraging agentic loops to massive mutual benefit:

 • Anthropic engineers use Claude Code to build Claude Code
 • Users use Claude Code to build their projects
 • Both loops feed the same flywheel

 The result? 100% of Claude Code is coded by Claude Code, creating an exponential development cycle:

 BUILD → SHIP → FEEDBACK → IMPROVE → ACCELERATE → (repeat faster)

This has led to unprecedented beneficial outcomes on both sides

 🔸 27x productivity multiplier (40,000 lines vs 1,500 for avg developer)   
 🔸 $1B to $7B run-rate in 9 months - fastest growing software co. ever
 🔸 "Foundational cohorts" that lock in, not churn out

OpenRouter calls it the "Glass Slipper" effect: when a tool perfectly addresses users' critical needs, users don't just adopt it - they integrate it so deeply into their workflows that they effectively lock and no longer behave like the typical AI "tourist" constantly searching for the next new and exciting thing

I believe those already focused on mastering the skills necessary to reap the benefits of this joint agentic loop are going to be big winners in 2026

 Swipe through for the full breakdown 👉

 ---
 #ClaudeCode #AI #AgenticAI #FutureOfWork #Anthropic #SoftwareEngineering

Experimenting at the edge of AI and data to make you a better analyst | 6x Linkedin Top Voice | Lifelong Data Geek | IBCS Certified Data Analyst

Claude Code's Revolutionary Producer-Consumer Shared Loop

Recently, I showed how my Claude Code #IBCS Chart Skill employs the same "agentic loop" I used to solve DAX, but with a visual target rather than a quantitative one.  Tonight, it goes up against the toughest #IBCS chart...

What i call the "Triple Chart" - (the IBCS – International Business Communication Standards  actual vs. previous year column chart with absolute and relative variances aligned above) has been the Waterloo of every attempt we've made over the past two years to develop a straightforward, accurate and reliable AI-based method for generating charts that conform to the IBCS standards.

I won't spoil the results of tonight's "heavyweight championship fight", but I will say that Adam Schroeder and the Plotly team, and Eszter Kovacs also have developed very promising approaches to meeting this same objective.  

In addition, Claude Code recently has released a number of powerful innovations aimed at more precise control of the agentic loop and the continual learning process. 

I believe that working together, we soon will crack this difficult nut, and have a flexible solution able to quickly, easily, and accurately generate a full array of ICBS charts.

#ai #claudecode #agenticloop #dataanalysis #data visualization

Jürgen Faisst
Rolf Hichert

P.S.  For those interested, here are links to my previous posts on using Claude Skills agentic loops for #IBCS and DAX:

https://lnkd.in/ek_6MKxP

https://lnkd.in/eH_iJFf4

taplio

The Battle

Leon Gordon

Brian Julius

Followers

Posts

Engagements

Likes

Performance ScoreLeon Gordon's LinkedIn Performance

Performance ScoreBrian Julius's LinkedIn Performance

Ready to grow your LinkedIn presence?

taplio

The Battle

Leon Gordon

Brian Julius

Followers

Posts

Engagements

Likes

Performance Score.css-goub6s{font-size:var(--chakra-fontSizes-md);font-weight:500;color:var(--chakra-colors-gray-600);display:block;margin-top:var(--chakra-space-1);}Leon Gordon's LinkedIn Performance

Performance ScoreBrian Julius's LinkedIn Performance

Ready to grow your LinkedIn presence?

Performance ScoreLeon Gordon's LinkedIn Performance