Lots of orgs are getting excited about LLM visibility and getting tools to help them monitor how they appear for various prompts. And, because everyone and their brother is interested in LLMs, i see way more people, from far wider across the organization, wanting access to these tools than anything we've brought online before. And for each person that wants access, 5 people are asking questions they have decided can be answered by the new tool. And some orgs are doing a really bad job of controlling who gets access at what permission level. Not to mention a bad job of prioritizing ad hoc asks. For most orgs the two following steps are working at least for day to day: - I've suggested people attend 20 minutes of training or watch a 20 minute video before they get access. - Ive drafted a tiny policy about what the tool is for (monitoring progress of llm visibility initiatives, not idle curiosity) and how to tag prompts you add/check before removing prompts. But ive got an org where everyone is too busy for the training and not sure if they really need a policy. Anyone have tips or recommended reading on how to sell people in this kind of situation on the need for governance?

One of the biggest data problems is not lack of data. It’s lack of understanding. When teams don’t know what data means, where it comes from, or how it should be used, trust breaks down and AI becomes risky. I’ve just added a new resource to the classroom: Metadata & Lineage Starter Map This template helps you: - clarify business definitions, - map data flows end to end, - reduce misinterpretation, - build explainable and reusable data foundations. You can access it now in the Data Governance from Scratch classroom. As always, start small and focus on impact.

Hey team! 👋 I am putting the final touches on the "CDMP Accelerator: Pass in 4 Weeks course". My goal is simple: turn this 600-page monster (the DMBOK) into a clear battle plan to get certified in under a month. But to make this surgical, I don't want to guess. I want to build exactly what you need right now. To help me calibrate the content (More basics? More exam traps? More strategy?), please tell me honestly where you are in your CDMP journey: 👇 Click the option that matches your current status:

Hello friends,I wanted to share a reflection sparked by a recent talk from Seun Ebenezer from Greencon on GenAI assisted data stewardship. One idea really stuck with me: GenAI in governance is no longer a future concept. It is already here. But its success depends less on the technology itself and much more on foundations like data quality, shared semantics, and human accountability. As we say in Spanish, you cannot build the house from the roof. No data. No AI. I would love to hear how others in this group are thinking about human in the loop models, AI ready data, and the shift from reactive to proactive governance. What is working in practice, and where are you still seeing friction? Looking forward to learning from your perspectives. https://lnkd.in/dgMk3cBg

Given the current state where application data is distributed across MongoDB (NoSQL/document store) and PostgreSQL (relational/SQL), we recommend a federated data governance model that respects each database's strengths while establishing unified governance controls. This approach balances centralized policy with decentralized implementation. 2. Guiding Principles Database-Agnostic Policies: Define governance requirements independent of underlying technology Proportional Controls: Implement controls based on data sensitivity, not database type Unified Metadata Layer: Single source of truth for data lineage, classification, and ownership Minimal Application Impact: Governance implementation should not require major application rewrites 3. Proposed Governance Framework 3.1 Organizational Structure Data Governance Council (Strategic) ↓ Data Stewards (Business Domain Owners) ↓ Database Custodians (Technical Implementation) ↓ Data Consumers (Application Teams)