Building Strong Foundations

Everyone's racing to build AI governance. Without checking the foundation. Organizations are standing up AI governance programs and committees even though their data layer is a mess. AI governance built on a weak foundation isn't governance. It's theater. The organizations getting this right aren't starting with models or vendors.  They're starting with data, with the practices most companies built years ago. Data governance isn't a legacy IT problem.  It's the foundation every AI capability is built on. In other words, both are critical. A good way to think about it is: 👉 𝗗𝗮𝘁𝗮 𝗚𝗼𝘃𝗲𝗿𝗻𝗮𝗻𝗰𝗲: 𝗣𝗿𝗼𝘁𝗲𝗰𝘁𝘀 𝗪𝗵𝗮𝘁 𝗚𝗼𝗲𝘀 𝗜𝗻 👉 𝗔𝗜 𝗚𝗼𝘃𝗲𝗿𝗻𝗮𝗻𝗰𝗲: 𝗣𝗿𝗼𝘁𝗲𝗰𝘁𝘀 𝗪𝗵𝗮𝘁 𝗖𝗼𝗺𝗲𝘀 𝗢𝘂𝘁 However, there's a lot of overlap between the two.  For example, here are 10 connections that often get overlooked: → Data quality management becomes model reliability → Lineage tracking becomes bias root-cause analysis → Access controls become ethical use boundaries → Data cataloging becomes model registry and discovery → Compliance frameworks become regulatory readiness → Data stewardship roles become model ownership → Schema standards become consistent training inputs → Versioning and change control become drift monitoring → Security and encryption become adversarial defense → Quality metrics become explainability requirements You can't build the structure without a strong foundation. The organizations struggling with AI governance aren't missing the right tools. They're missing the infrastructure that everything depends on. ♻️ Share if this resonates ➕ Follow Jason Moccia for more insights on AI and leadership. 👉 Sign up for my free newsletter to learn more: https://lnkd.in/enhk3j9c

Stop building defence startups without understanding defence markets. This may sound harsh, but it needs to be said. I meet many innovators and founders who want to “enter defence.” The ambition is there. The pitch decks are polished. But the strategy often isn’t there. Too many solutions are built with a single local market or own idea in mind. The assumption is that if something works in one country, it will naturally translate elsewhere. It rarely does. Defence markets are shaped by different doctrines, procurement systems, command structures, policies, industrial ecosystems, security requirements, and CULTURES. What resonates in one country can fall completely flat in another. Without understanding how to bridge those differences, market entry attempts almost always fall short. Another challenge is funding. Much of the funding available to early-stage defence innovators is still very local. Many investors have limited understanding of what building a defence capability for international markets actually requires. Questions often focus on familiar local startup metrics that simply do not apply. Defence is not a burger joint. It's a long game built on credibility, trust, integration, survivability in environments where failure has real consequences. Which leads to another recurring problem: positioning. “We build drones.” “We do AI.” “We provide autonomy.” That is not a capability description. Who is the operator? What mission problem are you solving? Where does the system integrate? How does it survive electronic warfare? How is it secured, governed, sustained, and supported? And most importantly: Why should anyone trust it when lives depend on it? Vagueness does not survive contact with defence reality. Commanders don't want snake oil. They want systems that work inside real operational ecosystems. That often means partnerships, integrations, logistics chains, doctrine alignment, years of credibility-building. This is where expectations frequently collide with reality. Access is not automatic. Trust is not immediate. Adoption does not happen in quarters. It takes years. What works today isn't automatically what another military is looking. Every environment has its own operational constraints, political context, integration requirements. Which means building for defence requires something many startup ecosystems underestimate: homework. Understanding the environment, the user, the system you are entering. Defence is never a one-way street. Collaboration is never a one-way street either. To receive access, insight, partnership, you also have to bring something meaningful to the table. The innovators who succeed are rarely the loudest but ones who listen first, ask the hard questions, spend time understanding operators, integration realities, mission constraints, the broader ecosystem before building the narrative. Because in defence, buzzwords don't matter. Survivability does. #DefenceInnovation #DefenceTech #NoBuzz

𝙎𝙬𝙞𝙩𝙯𝙚𝙧𝙡𝙖𝙣𝙙’𝙨 𝙏𝙖𝙨𝙠𝙛𝙤𝙧𝙘𝙚 𝘿𝙧𝙤𝙣𝙚𝙨 𝙅𝙪𝙨𝙩 𝙎𝙝𝙤𝙬𝙚𝙙 𝙃𝙤𝙬 𝙩𝙤 𝙒𝙧𝙞𝙩𝙚 𝘿𝙧𝙤𝙣𝙚 𝙍𝙚𝙦𝙪𝙞𝙧𝙚𝙢𝙚𝙣𝙩𝙨 🔍 One of the hardest problems in drone procurement isn’t “which platform?” It’s describing the mission in a way industry can build, test, and price—without turning it into doctrine. Switzerland’s Taskforce Drones did something refreshingly practical: standard operation scenarios + a menu of optional capabilities. 🎯 Four scenario “building blocks” (clear, testable, and procurement-friendly): ▪️ Attack drone: one-way precision strike against ground targets (incl. beyond line of sight), using external target detection. ▪️ Airdrop drone: UAV that releases an effector (not necessarily expendable like a one-way system). ▪️ Ambush: attack drone that can wait concealed on the ground at a choke point, then strike by surprise. ▪️ CUAS interceptor drone: very fast interceptor for kinetic engagement of small UAVs—cuing first, autonomous final approach second. 🧭 What makes it smart (and scalable): ▪️ Range brackets are explicit (e.g., up to 15 km, with optional steps beyond that). ▪️ Environment matters: rural/urban lowlands, low mountain range, high mountains. ▪️ Reality checks: day/night, rain, snow, storms, fog, wind—written as capability conditions, not excuses after trials. ⚙️ The procurement lesson: Stop buying “a drone.” Start buying mission packages with a minimum baseline, then add options like LEGO—effects, targets, ranges, environments. 𝘐𝘧 𝘺𝘰𝘶 𝘤𝘢𝘯’𝘵 𝘸𝘳𝘪𝘵𝘦 𝘪𝘵 𝘢𝘴 𝘢 𝘵𝘦𝘴𝘵𝘢𝘣𝘭𝘦 𝘴𝘤𝘦𝘯𝘢𝘳𝘪𝘰, 𝘺𝘰𝘶’𝘭𝘭 𝘱𝘢𝘺 𝘧𝘰𝘳 𝘪𝘵 𝘢𝘴 𝘢 𝘸𝘪𝘴𝘩. ✅ This is the kind of framing that accelerates market engagement, reduces ambiguity, and makes evaluation defensible. #DefenseInnovation #Drones #LoiteringMunitions #CounterUAS #MilitaryProcurement #RequirementsEngineering

RAPID CAPABILITIES OFFICES (RCO): How the DoD Delivers When Time Is the Enemy! Most defense programs take years—sometimes decades—to move from concept to capability. But what happens when we don’t have that kind of time? In JRAC, we often turn to the RCOs for an example of speed at scale. The Rapid Capabilities Offices (RCOs) are elite teams that operate across the Department of Defense to deliver critical technologies fast—often in months, not years. And they do it by rewriting the rules. Each RCO is a small, mission-driven unit with direct access to senior leadership and a singular goal: get warfighters what they need before the threat evolves. No endless PowerPoints. No multi-year delays. Just speed, focus, and execution. Examples *corrected*: • The Air Force RCO (DAF RCO) delivered the B-21 Raider bomber, leveraging advanced stealth and survivable C2. • The Army RCO, now part of the Rapid Capabilities and Critical Technologies Office (RCCTO), fast-tracked hypersonic and directed energy weapons. • Marine Corps RCO: Rapidly fielded Autonomous Low-Profile Vessel (ALPV)—a semi-submersible drone boat inspired by narco subs—to stealthily transport supplies or launch missiles. It’s now undergoing front line operational testing. • The Space RCO is fielding tactically responsive launch and resilient satellite constellations for the U.S. Space Force. These aren’t demo labs. They’re operational accelerators. They de-risk cutting-edge tech, prove it in real-world scenarios, and transition it into service programs at scale. So how do these RCOs fit into the bigger DoD picture? Think of them as spearpoints—complementing traditional acquisition systems by showing what’s possible when bureaucracy doesn’t get in the way. They partner with labs, Combatant Commands, and PEOs to translate innovation into impact. And increasingly collaborative with JRAC. If you’re a private sector company with a game-changing capability, here’s how to engage: 1. Align to the mission—RCOs aren’t looking for flashy tech, they’re looking for solutions to urgent warfighter problems. 2. Engage through the ecosystem—AFWERX, DIU, SpaceWERX, and other innovation hubs often serve as on-ramps. 3. Come ready—Classified work, rapid prototyping, and non-traditional contracts (like OTAs) are the norm. This model isn’t theoretical. It’s operational—and it’s helping the U.S. stay ahead in a world where our adversaries aren’t waiting around for a JROC brief. The bottom line? RCOs are what acquisition looks like when urgency, trust, and warfighter outcomes are in charge. Links follow. DAF RCO: https://lnkd.in/eS_tCVnF Space RCO: https://lnkd.in/eBsDNBrN Navy RCO: https://lnkd.in/ekzhvxeS USMC RCO: https://lnkd.in/e_arcFUF Army RCCTO: https://www.army.mil/rccto #RCO #RapidCapabilitiesOffice #JRAC #Defense #Innovation #Warfighter

Reference Data Frameworks - The Foundation You Didn’t Know You Needed Most organisations struggle with reference data because they treat it like “small data.” But reference data is high impact - it drives reporting, compliance, analytics, risk, customer experience… everything. To manage it properly, you need a Reference Data Framework - a simple, structured way to keep your business language consistent and controlled. Here’s what a strong framework includes: 1. Reference Data Strategy • Why reference data matters • Enterprise principles • Where it fits in your Data Governance model • How it supports CDEs, Master Data, and reporting 2. Clear Ownership & Accountability • Business owns definitions and values • IT owns platforms and enablement • Data Governance ensures standards and compliance 3. Standardised Reference Data Domains • Financial • Product • Customer • Risk • Operations Every domain has an approved list, definition, steward, and change path. 4. A Controlled Change Management Process • Single change intake • Assessment of impacts • Approval by the right data owner • Versioning and release No silent updates. No new values sneaking in. 5. Authoritative Source & Technology Layer • MDM or RDM tool as the single source • APIs to distribute values • No spreadsheets • Every system consuming the same truth 6. Data Quality Rules for Reference Data • Valid values only • No expired, duplicated, or unauthorised codes • Alignment with Master Data and CDE definitions 7. Monitoring, Metrics & Audits • Usage consistency • Cross-system alignment • Change volume • Issues and breakages Reference data should stay clean - not decay quietly. A reference data framework doesn’t just prevent errors. It stops operational risk, protects reporting, and keeps your business speaking one language. If you missed my last post on Reference Data vs Master Data vs CDEs, check it out - this framework builds directly on it. Watch out for my next post: “The 10 Biggest Reference Data Failures - and How to Avoid Them.”

AI won't fix your bad data. But a solid data foundation will transform your AI... Too many companies rush to implement AI before organizing their data. It's like building a skyscraper on quicksand. No structure. No consistency.  No strategy. This approach leads directly to: • Unreliable insights that mislead decision-makers • Inefficient AI models that waste computing resources • Thousands of dollars spent with minimal return The hard truth: Data is an ingredient. Intelligence is the outcome. You can't cook a gourmet meal with spoiled ingredients. (I haven't tried it but I'm guessing) A strong data roadmap solves these fundamental problems by: → Breaking down organizational silos → Structuring data for optimal use → Creating consistency across systems → Enabling truly intelligent decision-making Companies that invest in data structure will lead the AI revolution. The rest will struggle to keep up, constantly wondering why their AI investments aren't delivering. The difference isn't in the AI tools. It's in the data foundation. Our team at Michigan Software Labs addresses this head-on: 1. Data Discovery - Uncover what data exists and pinpoint any gaps. ~3 weeks. 2. Data Structuring - Organize and refine your data for clarity and quality 3. System Connectivity - Link platforms and tools to break down silos 4. AI Enablement - Apply AI solutions to well-prepared, structured data Stop throwing good money after bad. Start building the foundation your AI initiatives need to thrive. p.s. - If you've been following me for a while but we've never connected directly, I'd love to hear from you. Drop me a comment or send a quick note. Whatever professional challenge you're facing, I'm here to help - and if I can't, I’ll point you to someone who can.

AI First or Data First? Everyone is rushing to become "AI First." But there's a fundamental question we need to ask: What is AI built on? If the foundation is fragmented, inconsistent, and untrusted data, even the most advanced AI models will struggle to deliver reliable outcomes. Building AI on poor-quality data is like constructing a skyscraper on unstable ground. The real transformation starts with being Data First: ✅ Data Governance ✅ Data Lineage ✅ Clear Ownership ✅ Data Quality ✅ Meaningful Metrics ✅ Organizational Trust When these pillars are in place, AI becomes an accelerator of business value rather than a source of uncertainty. AI is not the foundation. Data is. Organizations that invest in trusted, governed, and high-quality data today will be the ones that realize the true potential of AI tomorrow. Data First → AI Success What are you seeing in your organization? Is the focus currently on AI adoption or on building a strong data foundation first?

Most organizations think AI transformation starts with models. It doesn’t. It starts with the data platform. Because AI is not just a technology layer added on top of the business. It changes how organizations operate, make decisions, automate workflows, and create value. And none of that works without a strong data foundation. An AI-ready data platform enables organizations to: – Connect fragmented data across systems – Create trusted and governed data foundations – Deliver real-time intelligence – Power AI models and agents at scale – Turn data into actionable outcomes Without that foundation, AI initiatives often become: – Isolated experiments – Disconnected copilots – Or impressive demos with limited business impact This is the shift many organizations are now realizing: AI transformation is not only about adopting AI models. It’s about redesigning the platform that powers intelligence across the enterprise. Because in an AI-native world, the data platform is no longer a backend system. It becomes the operational foundation for: – AI – Automation – Decision-making – And increasingly, autonomous agents The organizations that succeed with AI will not necessarily be the ones with the largest models… They will be the ones with the strongest data foundations. How is your organization approaching AI transformation today? Starting from the model layer… or from the data foundation itself?

Every transformation story looks bold on the surface—new markets, new products, new ambitions. But in my experience, growth almost always runs into a quieter constraint first: the foundations underneath it. As we scaled into new markets and worked with more sophisticated clients, it became clear that our biggest limiter wasn’t demand or vision—it was how data moved, scaled, and was reused across the enterprise. What worked well inside individual products wasn’t designed to scale together. In this article, I share why we deliberately went back to first principles—decoupling operational and analytical data, separating OLTP from OLAP, and investing early in a central data platform to create headroom for analytics, AI, and innovation without destabilizing core systems. This wasn’t about chasing new technology. It was about removing architectural constraints that quietly slow growth—and building a foundation that lets teams innovate in parallel, not in sequence. Would love to hear how others are thinking about data foundations as a growth enabler—not just a technical one. 👉 From Fragmented Data Operations to a Scalable Data Platform: Why We Chose to Go First Principles