entangled.work’s Post

📌 This post is part of a 3-part series on how AI is reshaping work—from intelligence on demand to human-agent teams and the rise of the agent boss. ————————————————————— Part 2. Signal #2: Human–agent teams will break the org chart. The traditional org chart was designed for a world where: • Expertise was scarce • Work was slow and sequential • Coordination happened through management layers That world no longer exists. When intelligence becomes on-demand, work no longer needs to be organized strictly by functions like marketing, finance, or engineering. Instead, it organizes around outcomes. This is where human–agent teams emerge. In these teams: • Humans define goals, constraints, and priorities • AI agents execute research, analysis, coordination, and follow-through • Teams assemble and dissolve based on the job—not the department The result is a shift from: ❌ Static hierarchies ➡️ Dynamic, goal-driven “work cells” Think less traditional corporate org chart. Think more project-based production, where the right mix of humans and agents comes together, delivers impact, and moves on. This doesn’t eliminate leadership or accountability. It changes where they live. Managers spend less time allocating tasks and more time: • Setting direction • Resolving ambiguity • Making judgment calls • Managing risk and exceptions Over time, the org chart won’t disappear overnight—but it will quietly lose relevance. What replaces it is a work chart: a living system that maps outcomes to the humans and agents best equipped to deliver them. The companies that adapt fastest won’t be the most automated. They’ll be the ones that restructure work to move at the speed of intelligence. 👉 Final post: why every employee is becoming an “agent boss.” ♻️ Repost if you found this valuable 🔔 Follow for real-world AI implementation in enterprise — beyond hype #FutureOfWork #OrgDesign #OrgTransformation #HumanAgentTeams #AIAtWork #Leadership #AgileOrganization #OperatingModel #WorkInnovation #DigitalTransformation #AITransformation #TeamDynamics

More and more reports are coming out every day, every week, every month, showing businesses failing with AI implementation. From my own experience in dealing with businesses and from these reports, there is one common issue. It's AI didn’t fail your business. Your business operating and data model did. Most “AI transformations” don’t die in the model. They die in the messy middle: • Unclear ownership (no one truly owns the workflow) • Broken workflows (copy/paste, approvals, workarounds) • Change fatigue (“another tool… That’s great… This is so cool... When will leadership stop this madness…?”) McKinsey and others are seeing 70% or more of large-scale transformations fail. How does a business not become part of the 70% plus failure rate? Use this sequence: Stabilize → Simplify → Automate → THEN AI 1. Can you map the workflow end-to-end in 15 minutes? 2. Do you trust your data definitions + system of record? 3. Can the team run the process without “hero” employees? When businesses work with Scale Crew, we help you design the operating model + optimize workflows and data first, then we automate, and finally (and only at the last step) do we let you begin discussing bringing in a custom-developed app or AI into the day-to-day. #ArtificialIntelligence #DigitalTransformation #Automation #ChangeManagement #WorkflowOptimization #TheScaleCrewHR

Yesterday someone said to me, “I’ve been reading your articles… but I still don’t know what AI-native means. I don’t even know what tool to pick.” And that’s exactly the point. AI-Native has nothing to do with choosing a tool. It has everything to do with choosing a mindset. I told her - Being AI-Native isn’t about knowing which tool. It’s about being open to the option of a tool and knowing where it belongs in the operating model. Because the modern Chief of Staff isn’t defined by software. They’re defined by their willingness to ask: “Could AI help us run this better?” “Where could it reduce friction?” “What part of this system is ready for intelligence, not effort?” The Chief of Staff doesn’t run tasks. They design operating systems. AI is simply one lever inside that system. 1. Shift the Mindset: From Doing → Designing You don’t need to be technical. You need to identify: bottlenecks friction points decision gaps places where AI can accelerate clarity Mindset is 90% of being AI-Native. 2. Build Operating Systems, Not Templates A modern CoS builds: decision flows interlocks operating rhythms governance KPIs accountability loops Tools don’t fix broken systems. Systems make tools powerful. 3. Use AI as a Force Multiplier AI should help you: summarize synthesize spot risks see patterns connect insights across teams If it takes hours, AI reduces it to minutes and as the CoS you know where to place it. 4. Ask Better Questions I told her: “The CoS who succeeds in 2026 isn’t the one who knows every tool. It is the one who knows what to ask.” Questions like: “What decision are we really making?” “Where does this break most often?” “Who needs visibility?” “Could AI eliminate this rework next time?” and most importantly " WHAT PROBLEM ARE WE TRYING TO SOLVE " This is the real work. 5. Build Confidence, Not Complexity Teams aren’t afraid of AI. They’re afraid of using it wrong. The CoS creates a safe operating model where experimentation is allowed and supported. Final Thought Being AI-Native isn’t about tech skills. It’s about curiosity, openness, and the confidence to redesign how the organization runs. If you can ask where AI adds value, reduce friction, and design smarter systems, you’re already becoming an AI-Native, Operational Excellence Chief of Staff. #ChiefOfStaff #OperationalExcellence #AINative #AILeadership #FutureOfWork #StrategyExecution #DigitalTransformation #OperatingModel #LeadershipDevelopment

✅ 𝗣𝗢𝗦𝗧 𝟮𝟴 — 𝗢𝗽𝗲𝗿𝗮𝘁𝗶𝗼𝗻𝗮𝗹𝗶𝘇𝗶𝗻𝗴 𝗔𝗜: 𝗗𝗲𝗽𝗹𝗼𝘆𝗶𝗻𝗴 𝗠𝗼𝗱𝗲𝗹𝘀 𝗜𝘀 𝘁𝗵𝗲 𝗦𝘁𝗮𝗿𝘁, 𝗡𝗼𝘁 𝘁𝗵𝗲 𝗙𝗶𝗻𝗶𝘀𝗵 One pattern I’ve noticed across many organizations is a sense of celebration when an AI model is deployed—as if the mission is complete. But every time I see a team pop the champagne at deployment, I quietly think: “𝗧𝗵𝗲 𝗿𝗲𝗮𝗹 𝘄𝗼𝗿𝗸 𝘀𝘁𝗮𝗿𝘁𝘀 𝗻𝗼𝘄”. Deployment isn’t the finish line; it’s the beginning of ownership. I’ve seen this movie countless times: a team deploys a strong model that performs well for months, but then something changes—user behavior, business strategy, or an upstream system. Suddenly, predictions drift, accuracy drops, users lose confidence, complaints increase, and everyone is shocked. But they shouldn’t be. 𝗔𝗜 𝗹𝗶𝘃𝗲𝘀 𝗶𝗻 𝗮 𝗺𝗼𝘃𝗶𝗻𝗴 𝗲𝗻𝘃𝗶𝗿𝗼𝗻𝗺𝗲𝗻𝘁; 𝗲𝘃𝗲𝗿𝘆𝘁𝗵𝗶𝗻𝗴 𝗮𝗿𝗼𝘂𝗻𝗱 𝘁𝗵𝗲 𝗺𝗼𝗱𝗲𝗹 𝗲𝘃𝗼𝗹𝘃𝗲𝘀. The teams that succeed treat 𝗔𝗜 𝗺𝗼𝗱𝗲𝗹𝘀 𝗹𝗶𝗸𝗲 𝗽𝗿𝗼𝗱𝘂𝗰𝘁𝘀, 𝗻𝗼𝘁 𝗽𝗿𝗼𝗷𝗲𝗰𝘁𝘀. They invest in a necessary operational framework: • Automated monitoring and drift detection. • Alerting systems and retraining pipelines. • Shadow mode testing and a clear rollback strategy. • Regular performance reviews with SMEs, product owners, and risk leads. I worked with one organization that implemented a monthly “model health check”—a simple 45-minute review that caught numerous issues early because the right people looked at the same indicators together. The biggest difference I’ve observed is mindset. When teams see a model as “done,” it decays. When they see it as “living,” it improves. Operationalizing AI isn’t glamorous, but it’s where real value is generated. Anyone can train a model; far fewer can keep it effective, reliable, and trusted over time. 👉 𝗜𝗻 𝘆𝗼𝘂𝗿 𝗲𝘅𝗽𝗲𝗿𝗶𝗲𝗻𝗰𝗲, 𝘄𝗵𝗮𝘁’𝘀 𝘁𝗵𝗲 𝗵𝗮𝗿𝗱𝗲𝘀𝘁 𝗽𝗮𝗿𝘁 𝗼𝗳 𝗺𝗮𝗶𝗻𝘁𝗮𝗶𝗻𝗶𝗻𝗴 𝗮 𝗺𝗼𝗱𝗲𝗹 𝗮𝗳𝘁𝗲𝗿 𝗱𝗲𝗽𝗹𝗼𝘆𝗺𝗲𝗻𝘁? #𝗠𝗟𝗢𝗽𝘀 #𝗔𝗜𝗔𝗱𝗼𝗽𝘁𝗶𝗼𝗻 #𝗘𝗻𝘁𝗲𝗿𝗽𝗿𝗶𝘀𝗲𝗔𝗜 #𝗗𝗮𝘁𝗮𝗦𝗰𝗶𝗲𝗻𝗰𝗲𝗢𝗽𝘀 #𝗔𝗜𝗣𝗿𝗼𝗷𝗲𝗰𝘁𝘀 𝗟𝗶𝗻𝗸 𝘁𝗼 𝗣𝗼𝘀𝘁27: https://lnkd.in/gYbNtWrN

𝗠𝗼𝘀𝘁 𝗽𝗿𝗼𝗳𝗲𝘀𝘀𝗶𝗼𝗻𝗮𝗹𝘀 𝗮𝗿𝗲 𝘂𝘀𝗶𝗻𝗴 𝗔𝗜 𝘄𝗿𝗼𝗻𝗴, 𝗮𝗻𝗱 𝘁𝗵𝗲 𝗱𝗮𝘁𝗮 𝗽𝗿𝗼𝘃𝗲𝘀 𝗶𝘁. I analyzed AI adoption across many teams and found something shocking: 𝗧𝗲𝗮𝗺𝘀 𝘂𝘀𝗶𝗻𝗴 𝗠𝗢𝗥𝗘 𝗔𝗜 𝘁𝗼𝗼𝗹𝘀 𝘄𝗲𝗿𝗲 𝗼𝗳𝘁𝗲𝗻 𝗟𝗘𝗦𝗦 𝗽𝗿𝗼𝗱𝘂𝗰𝘁𝗶𝘃𝗲 𝘁𝗵𝗮𝗻 𝘁𝗲𝗮𝗺𝘀 𝘂𝘀𝗶𝗻𝗴 𝗳𝗲𝘄𝗲𝗿. Here's what the top 10% do differently: 1️⃣ They redesign workflows BEFORE adding AI → 3x higher output gains vs. teams that just adopt tools → They ask, "What process needs fixing?" not "What tool should we buy?" 2️⃣ They use AI for repeatable thinking, not everything → Best for: Drafting, summarizing, data analysis → Still human-led: Creative work, strategic decisions, relationship building → Rule: If it requires judgment, keep it human 3️⃣ They measure outcomes, not activity → Wrong metric: "We used AI on 50 tasks today." → Right metric: "We cut report time from 4 hours to 45 minutes." → Quality + Speed > Number of tasks 𝗧𝗵𝗲 𝘂𝗻𝗰𝗼𝗺𝗳𝗼𝗿𝘁𝗮𝗯𝗹𝗲 𝘁𝗿𝘂𝘁𝗵: If you're adding AI without changing how you work, you're just doing bad processes faster. My prediction for 2026: Companies will split into two camps: Those who use 𝗔𝗜 𝘁𝗼 𝗲𝗹𝗶𝗺𝗶𝗻𝗮𝘁𝗲 𝘄𝗼𝗿𝗸,  and  Those who 𝘂𝘀𝗲 𝗶𝘁 𝘁𝗼 𝗱𝗼 𝗺𝗼𝗿𝗲 𝘄𝗼𝗿𝗸. One will win. One will burn out. 𝗪𝗵𝗶𝗰𝗵 𝗰𝗮𝗺𝗽 𝗶𝘀 𝘆𝗼𝘂𝗿 𝘁𝗲𝗮𝗺 𝗶𝗻? 👇 #AIProductivity #FutureOfWork #WorkSmarter #BusinessAnalysis #DataDriven #ProfessionalGrowth #Leadership

𝗧𝗵𝗮𝘁'𝘀 𝗮 𝗪𝗿𝗮𝗽 𝗼𝗻 𝟮𝟬𝟮𝟱. 𝗜𝘀 𝗬𝗼𝘂𝗿 𝗧𝗲𝗮𝗺 𝗥𝗲𝗮𝗱𝘆 𝗳𝗼𝗿 𝘁𝗵𝗲𝗶𝗿 𝟮𝟬𝟮𝟲 𝗡𝗲𝘄 𝗔𝗜 "𝗩𝗶𝗿𝘁𝘂𝗮𝗹 𝗖𝗼-𝘄𝗼𝗿𝗸𝗲𝗿𝘀"? 🚀 McKinsey research suggests that by 2030, $2.9 trillion in value could be unlocked—but only if organisations prepare their people and redesign their workflows. The reality? AI adoption often takes much longer than necessary because human skills and organisational cultures adapt slowly. In 2026, Human-in-the-Loop (HITL) will be the default. We must stop treating AI as a software rollout and start treating it as a People Operations transformation. 𝟱 𝗞𝗲𝘆 𝗧𝗿𝗲𝗻𝗱𝘀 𝘁𝗼 𝗪𝗮𝘁𝗰𝗵: 𝟭. The AI fluency surge has increased the demand for managing AI tools by 7x in just two years. It is no longer a "nice-to-have". AI Fluency is now a leadership requirement. 𝟮. High-stakes verification, which demands greater cognitive effort, requires humans to take responsibility for collaboration, intent, emotion, and ethics—nuances that AI cannot yet replicate. 𝟯. Precision coaching is more effective than generic training for broad reskilling, which remains a half-hearted strategy. We need diagnostic intelligence to pinpoint the critical 20% of human actions that will generate 80% of AI-enhanced performance. 𝟰. Inference-time reasoning means we are moving toward systems that "think" and plan before they act. The human role is no longer to fix the logic, but to judge the strategy and amplify their expertise without needing to think like engineers. 𝟱. We must de-risk the 'Organisation Brake'—the silent buildup of structural friction and inefficient workflows that hinder progress and cost businesses thousands in lost productivity. Just as a high-performance engine cannot run with the handbrake engaged, advanced AI agents cannot perform if the underlying human operating system is unstable. Releasing this brake is crucial to drive the true potential of a new workforce. 𝗜 𝘀𝗽𝗲𝗰𝗶𝗮𝗹𝗶𝘀𝗲 𝗶𝗻 𝗔𝗜 𝗲𝗻𝗮𝗯𝗹𝗲𝗺𝗲𝗻𝘁 𝘀𝘁𝗿𝗮𝘁𝗲𝗴𝗶𝗲𝘀 𝘁𝗵𝗮𝘁 𝘁𝘂𝗿𝗻 𝗵𝘂𝗺𝗮𝗻-𝗳𝗮𝗰𝘁𝗼𝗿 𝗿𝗶𝘀𝗸𝘀 𝗶𝗻𝘁𝗼 𝗾𝘂𝗮𝗻𝘁𝗶𝗳𝗶𝗮𝗯𝗹𝗲 𝗯𝘂𝘀𝗶𝗻𝗲𝘀𝘀 𝗼𝘂𝘁𝗰𝗼𝗺𝗲𝘀. Let's talk about how to build this connective structure into your organisation. 👉 DM me to chat 📩

From Craft to Factory (Without Becoming a Sweatshop) Every services organization starts as a craft. A small group of talented people. High touch. High pride. High variability. As demand grows, the fear sets in: “If we standardize too much, we’ll lose what makes us special.” The truth is the opposite. What breaks teams at scale isn’t structure. It’s pressure without design. When growth outpaces operating discipline: • Utilization becomes erratic • Onboarding is inconsistent • Quality varies by team or region • Leaders compensate with heroics Well-designed scale looks different: • Repeatable onboarding and ramp plans • Clear delivery roles and expectations • Utilization targets that protect people and margins • Field enablement that reduces friction instead of adding process This is also where AI automation actually works. AI doesn’t fix broken services. It amplifies well-designed ones. When services are clearly defined, AI can: • Automate SOW creation and scoping • Accelerate planning, scheduling, and reporting • Surface delivery risk earlier • Reduce low-value administrative load on teams When services aren’t designed, AI just accelerates chaos. The best services organizations use AI on top of: • Productized services • Standard delivery models • Clear data and process ownership Craft doesn��t disappear at scale. It moves upstream—into service design, innovation, and continuous improvement. If growth feels exhausting, it’s not because you’re scaling. It’s because you’re scaling without intent. Design the system. Automate the repeatable. Protect the people. #ITServices #ProfessionalServices #Leadership #AI

We're still doing M&E like it's 2005. I've spent years in development work, and here's what I see happening across organisations, large and small: We're drowning in Excel/Google Sheets. We're waiting weeks for reports that should take hours. We're asking field teams to fill out the same forms manually, over and over again. And by the time we get the data, analyse it, and share insights... the program cycle has already moved on. This is the reality: Traditional M&E is too slow, too manual, and too disconnected from decision-making. We have the tools to fix this! AI and automation aren't just buzzwords. They're practical solutions that could transform how we work. Yet most organisations are barely scratching the surface. Imagine this instead: Real-time dashboards that update automatically as data comes in, AI that spots patterns and flags issues before they become crises, Chatbots that help field staff submit accurate data in minutes, not hours, Predictive models that help us course-correct mid-program, not after it's over This is what M&E should look like in 2026. So what needs to change? First, we need to stop treating M&E as a compliance exercise and start treating it as a strategic function. If we want real-time insights, we need real-time systems. Second, we need to embrace automation as a way to free ourselves from tedious tasks so we can focus on what actually matters: interpreting data, telling stories, driving impact. AI is not a threat to our jobs. Third, we need to upskill. AI literacy isn't optional anymore for M&E professionals. It's essential. The development sector talks a lot about innovation. But when it comes to how we measure and learn? We're stuck. And the cost isn't just inefficiency, it's missed opportunities to serve communities better, faster, and smarter. It's time to move M&E from the back office to the front lines of impact strategy. If you're working in monitoring, evaluation, or learning, or if you're leading programs that depend on good data, I'd love to hear from you. What's holding your organisation back from modernising M&E? What's one thing you'd automate tomorrow if you could? Let's start this conversation.

We are automating tasks at record speed. But we haven’t paused once to redesign how work should actually happen. That’s the hidden problem no one is talking about. AI is writing emails, generating reports, creating code, scheduling meetings, and handling support tickets. On paper, productivity looks higher than ever. But in reality, many teams feel more exhausted than before. Why? Because automation is being added on top of broken workflows. We’re automating steps that shouldn’t exist. We’re speeding up processes that were poorly designed in the first place. We’re asking AI to do more work instead of asking why the work exists at all. So instead of fewer tasks, people now manage: 📌 more outputs 📌 more notifications 📌 more tools 📌 more decisions Automation without redesign doesn’t reduce work. It amplifies chaos. Real progress doesn’t come from doing the same work faster. It comes from rethinking the work itself. Before automating, we should be asking: 📌 Can this step be removed? 📌 Does this task still make sense? 📌 Is this outcome actually valuable? 📌 Who should really be responsible for this decision? The companies winning with AI are not the ones with the most tools. They are the ones redesigning roles, responsibilities, and workflows from scratch. They don’t automate tasks. They redesign systems. The future of work is not about speed. It’s about clarity. And until we redesign work itself, automation will keep making us busy — not better. If you want, I can share a simple framework to redesign workflows before automating them with AI. Comment “WORK” and I’ll share it. #FutureOfWork #AI #Automation #Productivity #Leadership #WorkDesign #DigitalTransformation #Innovation

9 in 10 enterprise leaders just confirmed something big. (AI agents aren't replacing work—they're redesigning it) Anthropic surveyed 500+ technical leaders on AI agent deployment. And beyond all the productivity metrics and ROI numbers, there's a deeper shift happening. Here's what employees are now doing MORE of: → Strategic activities → Relationship building → Skill development And LESS of: → Routine execution → Manual processes → Repetitive tasks This isn't about replacing people. It's about elevating what people actually do. The numbers tell the story: 57% of organizations deploy agents for multi-stage workflows. 16% run cross-functional processes across multiple teams. By 2026? 81% plan even more complex implementations. What this looks like in real businesses: At Doctolib: Engineers aren't writing less code. They're shipping features 40% faster and tackling bigger problems. At eSentire: Security analysts aren't doing less analysis. They're reviewing AI-compressed insights (7 minutes vs 5 hours) and focusing on strategic decisions. At Thomson Reuters: Lawyers aren't practicing less law. They're accessing 150 years of case law instantly and spending time on client strategy. The reality check: Only 3 things are holding organizations back: 1. Integration with existing systems (46%) 2. Data access and quality (42%) 3. Change management (39%) Notice what's #3? Change management. Because this shift in how people work requires new skills, different workflows, and team adaptation. My observation: The organizations winning with AI agents aren't just deploying technology. They're redesigning work itself. The 2026 question isn't "can we deploy agents?" It's "how do we help our teams work differently?" Because technology is the easy part. Changing how people work? That's where the real challenge lives. P.S. How is AI changing the type of work your team focuses on? Drop real examples below—I'm genuinely curious 👇 #ai #aiagents #llm