Andreas Welsch's Linkedin Analytics

Three cities, three perspective on the state of AI adoption—and what’s needed next. Over the past three weeks, I’ve attended three industry conferences that brought together a cross-section of AI/ Technology, Data, and HR leaders. Needless to say, AI was front and center, but every time through a different lens. In tomorrow’s issue of The AI MEMO, I'll share a snapshot of the key topics and what leaders should focus on next to drive AI adoption. What to expect: Philadelphia — Philly Tech Week Insights from the “Leading AI Innovations” panel hosted by DataPhilly, with more than 200 attendees, including dozens of IT & Data leaders and AI professionals. Boston — Data Summit Key tips for “getting your business AI ready” and overcoming AI resistance from my keynote, and the top questions from Heads of Data & Analytics to Data Governance and Data Scientists building the foundations of AI products. Washington, D.C. — Association for Talent Development Top-of-mind topics for HR leaders supporting AI adoption through leadership development and upskilling. Conversations with 200 attendees reveal a broad spectrum of AI & workforce readiness. Subscribe to The AI MEMO today to get the next article in your inbox tomorrow and learn how AI, IT, and HR leaders are driving AI adoption: https://lnkd.in/g9BgxaXZ #ArtificialIntelligence #GenerativeAI #Leadership

An AI leader asked me how to improve the collaboration with business stakeholders: “I told them: This is what we do and this what we need from you. How we get more buy-in?” I wasn’t too surprised that “the business” has been apprehensive if the tone has indeed been “us vs. them.” That’s why I suggested to reframe and rephrase it: “Our team understands the technology really well and we are looking to partner with you to uncover the most promising AI opportunities together, based on your domain expertise.” But that is just the first step. Next, we talked about governance. Creating a simple table of roles & responsibilities can already increase transparency and drive alignment (AI team | Business team). Add the technical and business roles you need to bring together to work on an AI idea, consider what each role brings to the project, and who meets with whom and how often. Building on that, we talked about the typical project phases from idea to operation to show the project flow. Add the deliverables and documents needed for each phase along with the outcomes and go/no-go criteria for the project. (Check out the chapter on building your idea funnel in the AI Leadership Handbook.) Lastly, we covered getting sign-off on this governance framework across your senior business stakeholders. This will set you up for an aligned approach with top-down support and help you shine in your AI leadership role. I’ll check in again in a few weeks and can’t wait to hear how things are going. What’s slowing your AI program down? (Drop me a DM for an unbiased perspective.) #ArtificialIntelligence #GenerativeAI #IntelligenceBriefing

There's a huge difference between the simplicity you see in YouTube tutorials and what it takes to bring Agentic AI into your business environment. Expectations are high. The question "What's our AI strategy?" has quickly evolved into "What is our Agentic AI strategy?" for many IT organizations. But a push to "do something" with agents is neither easy nor automatically the right one. Historically grown landscapes and systems, heterogenous environments, data silos, homegrown apps, monolithic architectures, and everything in between, make it hard to innovate and add AI on top. Instead, look toward building composable architectures and services that can be exposed to AI agents more easily. This enables your team to build with (and on top of) your current applications and make key information in your business systems available to your end-users and customers. For example, start by building an agentic workflow that can handle more uncertainty and variance in the decisions than strict "if-this-then-that" ones—and expand it step-by-step with your company’s key applications. That’s what safe, secure, and scalable enterprise-grade integration is for. MuleSoft has a great resource to learn more and get started: https://lnkd.in/devCnTmr Next Wednesday, I’ll be speaking with MuleSoft’s executives in New York. Drop your questions in the comments and I’ll ask the top 3. #AgenticAI #MuleSoft #SalesforcePartner

Anyone will become a leader. Working with AI agents we ask them to create an outcome rather than follow defined steps. But this key task takes skill and practice: delegating. I was fortunate to participate in several leadership trainings in my career: as an aspiring leader, a first-level leader, and a mid-level leader. While each offered relevant insights for the role at the time, one thing has stayed consistent: the art of delegating a task. Delegate in passing and you’re leaving too much ambiguity to fill. Delegate to someone who’s missing the context and it will show in the result. But delegating starts with you: the leader. This simple framework can help you articulate the goal to be achieved more clearly and thereby set up your team member for success. Coincidentally, you can apply the same steps when delegating to AI: 1) What is the GOAL you want to achieve? 2) What DATA or information is needed to complete the task? 3) What TOOLS are available for the task? 4) How will you EVALUATE the output? (What does 'good' look like?) 5) With whom should they COLLABORATE? And lastly, asking what they have understood. Next time you delegate a task to a person or to AI, try structuring your request using these five steps. What would you add? #ArtificialIntelligence #GenerativeAI #IntelligenceBriefing

Will AI eliminate entry-level roles? A student recently asked me: "I'm about to graduate. Will I still find a job—because of AI?" It's a question that deserves a thoughtful answer. While Anthropic's CEO predicts AI could replace half of all entry-level roles within five years, and companies like Moderna are already eliminating junior positions, I believe we're looking at job evolution rather than elimination. Instead of entry-level roles disappearing entirely, I see expectations shifting upward. What we consider "entry-level" today will require higher proficiency tomorrow. Recent graduates will need to deliver work quality that previously required several years of experience, but they'll have AI as their co-pilot. This creates a ripple effect: entry-level workers perform at professional levels, professionals at expert levels, and so on. Companies are slowing entry-level hiring right now. But many will likely discover that roles can only be partially automated and humans are still essential. We're seeing contraction, not complete elimination. Students and recent graduates need to build human networks, acquire domain knowledge, and gain practical experience with AI tools. But that’s often a Catch 22. Educational institutions and businesses must bridge the gap by evolving how they teach and evaluate capabilities. What's your take on preparing the next generation for an AI-integrated workplace? Read the full article in my newsletter, “The AI MEMO”: https://lnkd.in/dqG9-zb5 #ArtificialIntelligence #GenerativeAI #IntelligenceBriefing

Are you falling for the AI hype? Amid the talk and claims in the market, it’s important to distinguish between truly agentic AI and ones that simply automate predefined workflows. While "agentic" is the buzzword of the moment, many vendors are actually delivering solutions that mimic past technologies rather than offering genuine decision-making autonomy. Workflows are still defined by humans while agents take on more complex and less well-defined goals to work independently. Engaging directly with these tools can help you identify which solutions genuinely provide value, rather than being swept up in inflated claims. Instead of rushing to adopt the newest technologies, look for use cases demonstrating real success, such as research or coding. This article in my weekly newsletter AI MEMO emphasizes the early stage of agentic AI development and the necessity for personal experimentation. By grounding your understanding in reality and testing these technologies, you’ll develop a more informed perspective that empowers your leadership. #ArtificialIntelligence #GenerativeAI #IntelligenceBriefing

Leaders are talking about transforming their business into an “AI-first” company, only to walk it back after getting backlash. What’s needed more than being first is being ready…to become first. The recent Axios survey shows that the majority of Americans wants AI companies to take a slower and more deliberate approach instead of an AI race—and that is across all age groups! But running successful AI programs in a business takes more than a top-down memo or mandate. Across 90+ interviews with AI leaders, I found these 9 principles apply: 1) Be aware of industry and tech trends 2) Align business and AI strategies 3) Establish a central AI leadership role 4) Design for people with a product-centric view 5) Cultivate an AI-ready culture 6) Harness diverse perspectives 7) Establish principles for human-centered AI 8) Ensure trusted, relevant results 9) Mitigate data privacy and security risks At Machine Learning Week in Phoenix next week, I’ll share the details on each of these aspects and how to turn technology hype into business outcomes. Who’s attending in person? #ArtificialIntelligence #GenerativeAI #MLWeek

𝗔𝗿𝗲 𝘆𝗼𝘂 𝗿𝗲𝗮𝗱𝘆 𝘁𝗼 𝗹𝗲𝗮𝗱 𝘆𝗼𝘂𝗿 𝗼𝗿𝗴𝗮𝗻𝗶𝘇𝗮𝘁𝗶𝗼𝗻 𝘁𝗵𝗿𝗼𝘂𝗴𝗵 𝗔𝗜-𝗱𝗿𝗶𝘃𝗲𝗻 𝗰𝗵𝗮𝗻𝗴𝗲? There's a lot of hype and fear in the headlines and it's starting to influence people's thinking. Amid all the talk about AI and disruption, adopting a human-centered mindset for AI adoption is important. In my latest newsletter, I share the strategies leaders can employ to navigate this change effectively. Understanding that AI is not merely a technology but a transformation of organizational culture is crucial. When leaders prioritize transparent communication and foster an environment of trust, they can demystify AI for their teams and successfully integrate it into the workplace. Leaders must also actively engage with their teams to identify ongoing AI initiatives and create opportunities for hands-on experience rather than jumping straight into upskilling. Training without direct application creates more fear than it helps. AI can enhance human capability, but this journey requires collective effort. This is just an excerpt from my weekly newsletter, in which I share insights and actionable strategies for business leaders adopting AI. Subscribe today to read the full issue and elevate your leadership skills with the next article tomorrow. #ArtificialIntelligence #GenerativeAI #IntelligenceBriefing

CEOs proclaiming their company will become “AI first” does not matter, until and unless the company can become “AI ready” first. To do that, leaders need to provide three things: 1) Access to relevant tools 2) Encouragement to use them 3) Hands-on training to increase AI literacy And, by the way: No, Copilot (alone) is not your AI strategy. This is a snippet is from my keynote at #DataSummit in Boston. Drop me a DM to get your business “AI ready.” What are you seeing? Tools, encouragement, training? #ArtificialIntelligence #GenerativeAI #IntelligenceBriefing

Leaders I speak with feel overwhelmed trying to stay on top of AI news. They ask: How do I know what’s actually worth paying attention to? Here’s the approach I share: Don’t chase every headline. Instead, look for patterns. If a topic is still showing up in the tech or industry news and in executive briefings six weeks from now, it’s probably meaningful. That’s when you dig deeper to evaluate the business potential. Yes, that's how short the cycles in AI are. Leaders don’t need to follow everything, but they need to apply a strategic filter. In my keynotes, I walk through a framework to help leaders identify trends that matter, map them to opportunities, and take informed action. If your organization needs a practical approach to tracking AI developments and translating them into business value, I’m available to present this in a leadership session or conference keynote. How do you filter signal from noise when it comes to AI trends? #ArtificialIntelligence #GenerativeAI #IntelligenceBriefing

Do you still need to disclose your use of ChatGPT or Copilot? A recent study by Duke University confirms what Slack discovered last fall: Employees are concealing their use of AI in the workplace because they are afraid to be seen as lazy or incompetent. But the actual problem is not your use of AI. It’s the perception that someone has taken a shortcut and left diligence by the wayside. It’s similar to the 2010s’ email signature “Sent from my iPhone (excuse any typos).” Whether or not there are any typos is less an issue of the device as it is of your process to skip proofreading the message before pressing Send. What if you’ve worked with a junior team member on the project? What if you’ve received feedback from a more experienced colleague? Would you disclose the full list of people? Credit where credit is due. But you don’t need to explicitly share that “Chat” or Copilot have been your sidekick on this project or email. Just do it. Permission granted. But don’t skip reviewing and understanding the output before sending it: 1) A simple reply to an email. Skim for tone and accuracy to ensure it sounds like you would write and that there are no obvious mistakes in it. 2) A market research report involving multiple sources and documents, hypotheses, and recommendations. At a minimum, read it top to bottom to check for obvious hallucinations, spot-check a few sources to check accuracy, look for inconsistencies in the information and conclusions. 3) A submission to an academic journal. Check the guidelines and apply the same rigor as if a co-author had written parts of it. This is 2025, not 2015. So, it’s time we normalize using AI instead of stigmatizing it. I’m working on two new courses with LinkedIn Learning to help managers and professionals navigate these dynamics in an AI-enabled workplace. I’d love to hear from you: Do you tell others you’re using AI and why (or why not)? ——— Sent from my iPhone 4 (excuse any typos) #ArtificialIntelligence #GenerativeAI #IntelligenceBriefing

How do you know that AI Agents are moving beyond exploration? Vendors like MuleSoft are adding support for Agent-to-Agent (A2A) communication and Model Context Protocol (MCP). What does that mean? Imagine you want to submit a purchase order via your AI assistant: 1) Using MuleSoft Anypoint Studio or Anypoint Code Builder, you create an MCP Server. 2) Create tools that enable your MCP Server to access live data (e.g., to your procurement application). You can also use your existing APIs. 3) Make the MCP Server available in your enterprise AI assistant (e.g., Claude). 4) Test it to ensure it works as intended. Then, test it again... 5) Use it in production to gather data on open purchase orders or to create new ones.. Scale it from a pilot group to additional teams and units while gathering feedback. You can even take it a step further and extend it by allowing agents to communicate with other agents (via A2A) to build out your agentic AI team of experts. Check out the full details: https://lnkd.in/davphiE2 I'm looking forward to exploring MuleSoft's MCP and A2A integration at the Connect AI event in New York City on June 25. Stay tuned for more... #AgenticAI #MuleSoft #Sponsored

31% of employees sabotage their company’s AI adoption. It’s hardly because of the technology. Your leaders want you turn AI hype into business outcomes. But it’s the first time your company goes through that kind of monumental change. It’s not just a matter of picking a vendor and running with it—or you’ll run into a dead-end quickly. Based on more than 60 interviews with AI leaders, you’ll discover how to: - Leverage the full scope of an AI leadership role - Win (enthusiastic) buy-in from employees - Take a product-centric approach to building AI applications - Build a pipeline of high-value AI capabilities - Utilize AI ethically, safely, and sustainably - Prepare your organization for agentic AI Get a head-start with the AI Leadership Handbook audiobook this weekend! #ArtificialIntelligence #GenerativeAI #IntelligenceBriefing

Yesterday, an AI leader asked me: “Should we keep our Center of Excellence centralized or start moving data scientists into the business?” It depends. That’s why I’m sharing a few tips to help you answer that strategic question: What’s the size of your company and your AI team? - If your company has 100-500 employees, a centralized CoE will likely be more beneficial. You can set guidelines, standards, technologies, and processes centrally without the overhead of working across business functions (or influencing). What roles do you have on the AI team? - If your team is small (think: “two pizza teams”) and builds AI products data preparation to application development, keep it as one to keep speed, momentum, and your ability to execute. Which business units do you engage with the most?/ Where do you create the most value? - If your highest number of AI projects comes from Procurement and your CoE is no longer able to help other business functions, consider adding headcount to your AI team first. If the projects (and value!) continue to increase, consider moving resources into the business or explore if the Procurement team might need its own satellite AI team (tat stays connected to your CoE). How are you getting ideas and signals from these business units?/ To what extent can tech-savvy multipliers be the connection? - Consider creating a community of multipliers in which stakeholders from the business come together to learn about what AI can do and what tools are available in your company. These multipliers become the go-to person for anything AI in their function and serve as idea scouts that help you vet ideas before they reach your CoE. In yesterday’s conversation, we concluded that keeping AI central and expanding the multiplier community is the best approach for the company at this time. But now I’m curious: How is your AI team organized? PS: If you weren’t at Machine Learning Week this week to talk in person, you’ll find a whole chapter in the AI Leadership Handbook on CoEs and their evolution as well as on how to build a multiplier community. #ArtificialIntelligence #GenerativeAI #IntelligenceBriefing

Well, that escalated quickly, no? For the last two years, we’ve heard phrases like: “AI won’t replace you—a person using AI will.” (Whether or not these phrases are actually more harmful than helpful is a different story.) “Do you think we’ll still find a job when we graduate?” one of my students asked this spring when we discussed AI in class. “Will we still have a job when all of this AI comes?” a sales leader asked me this week. “Our role is going to change a lot when AI can do all of that, and I’m pretty scared,” a marketing executive shared with me, referring to Meta’s AI-driven ad creation and media buying. Yet, when I asked whether and how they use AI for work, this future is unevenly distributed. While many smaller firms are figuring out what to do with AI (aka which tools to use and how to prompt them), the possibilities are already at least one generation ahead of that with AI agents working (semi-)autonomously. Some of this change is indeed happening over night. But innovation diffusion traditionally doesn’t happen all at once. So, there is ‘some’ time to figure this out individually. I expect that over the next few quarters and years, we will hear more questions like the ones above and more examples from our friends and families where change is happening. The question is just whether is happening with us, to us, or because of us. What do you think? #ArtificialIntelligence #GenerativeAI #IntelligenceBriefing

ICYMI: 👇🏻 Thank you, SwissCognitive for inviting me to join the Global AI Ambassador program for the fourth year in a row. The connections with fellow AI leaders from across the globe and the insights to bring AI into the lives of professionals and consumers make this a very special group.

I was throughly impressed! While generative and agentic AI are dominating the headlines, Machine Learning Week in Phoenix was a great reminder that we must not dismiss proven, predictive methods for decision making. For example: Reducing the number and value of retail fraud, predictive models detect suspicious transcations. Generative AI helps harmonize data and generate ready-to-send reports to the authorities. Increasing online subscriptions by using a predictive model that calculations the probability of a user subscribing if they are shown a paywall on an article. There are hard dollars and ROI behind these models, and the key in all these scenarios is knowing which technology to use when. In additional sessions, leaders from Paychex and others shared best practices on building a community of practice that increases the view of data and processes across the organization, and essential learnings from running AI projects such as "start with the business question before writing your first line of code." In my talk, I shared the 9 steps of successful AI programs that AI leaders need to keep in balance simultaneously. If you've missed it, you can also find them in my AI Leadership Handbook or invite me to share them in a session with your team. It was great to see a dedicated focus on Machine Learning and predictive modeling, and networking with Data and ML leaders who are doing the real work day in and day out. Predictive, Data Science, ML, AI—What term do you most identify with and why? PS: Thank you, Eric Siegel & team, for inviting me to speak in Phoenix. And as Eric shared with me: Half the sessions were actually about GenAI. I guess, I subconsciously attended the ones that were not... :) #ArtificialIntelligence #MachineLearning #MLWeek

Half of all entry-level roles could vanish, thanks to AI. It's a big claim. I see a common misconception when pondering what that means and how it will unfold. It is assuming that the scope of entry-level roles remains static while the majority of these tasks will be delegated to AI (agents). Instead, I believe we will see a shift in what “entry-level” means as well as what “professional” experience entails. I visualize this as an upward shift of the tasks that (a) will be deemed to be entry-level, (b) will be done by humans, and (c) the point in time at which this will occur. This shift won't happen overnight. But C-suites will drive this as a cost and efficiency play. The first signs are there: Klarna, Shopify, Duolingo, Moderna, etc. This also means that expectations towards entry-level applicants will change as well. Recent graduates will be able to create work products and quality with the help of AI that a professional with a few years of experience has been delivering without AI. Likewise, professionals can perform at an expert level. And so on. Similar to a sliding window, the scope of the role will shift towards combining domain expertise plus how to use AI to create "good" results. The question is how organizations and education providers can adapt (and how quickly) to prepare the workforce. Tectonic shift or "just another day in the office"? What do you think? #ArtificialIntelligence #AgenticAI #EntryLevel

I hit a LinkedIn milestone last week: 30,000 followers! Time to re-introduce myself to those of you who’ve recently started following me: Hi, I’m Andreas. After 23 years in senior AI leadership roles in big tech (at SAP), I started my own business last summer. My mission is to help new and aspiring AI leaders successfully implement AI in business, creating the resources that I wish existed when I was going through it. Working on AI with the largest brands on the planet for the past 10 years, I share my insights building on four perspectives: 1) Business Development & Go-to-Market -> Assessing an organization’s “AI-readiness” -> Selecting the right technology for the task (and how much to charge/ pay for it) -> Assembling a nimble team and setting an AI vision & strategy 2) Product Management -> Creating a portfolio of valuable AI ideas -> Scaling an AI program via a multiplier community -> Influencing business stakeholders across the company 3) Product Marketing -> Standing out in the market when everyone is shouting “AI!” -> Understanding business leaders’ needs and speaking their language -> Creating flexible monetization models that scale 4) Teaching/ Upskilling & Speaking -> Contextualizing industry trends & shifts -> Making complex concepts simple -> Getting a room full of strangers excited and keeping them engaged -> Bringing your most important asset along on the journey: people As an AI leader, you need to: - identify those AI ideas that create value (not just feel-good vibes), - scale your initiative across the company (not get stuck in your silo), - advocate for your team with clarity (not superlatives and jargon), and enable everyone in your company to use AI (not just technologists). What else would you add? Thank you for all the connections and discussions! PS: If you’re looking for help on these topics, DM me to plan a keynote, workshop, or campaign. - - - - Here are a few of my resources to help you grow your AI leadership: - Book: AI Leadership Handbook—A Practical Guide to Turning Technology Hype into Business Outcomes - Live Stream & Podcast: What’s the BUZZ?—AI in Business - Newsletter: The AI MEMO - Courses: LinkedIn Learning — Agentic AI, Managing Risks in AI Projects (upcoming) #ArtificialIntelligence #GenerativeAI #IntelligenceBriefing

Since 2022, SwissCognitive has brought together top AI experts and thought leaders as part of the Global AI Ambassador program. The mission is simple: to elevate awareness for AI without borders. This year’s group of ambassadors, once again, represents a diverse range of backgrounds. For example: - Jeff Winter is one of the leading thought leaders on AI in Manufacturing - Johan Steyn explores inclusive AI for businesses in South Africa - Michael Wegmüller & team help businesses in Switzerland be at the forefront of innovation Be sure to follow them as well as the other ambassadors to bring some fresh perspectives into your feed beyond the top cheat sheets for ChatGPT & Co. Thank you, SwissCognitive | AI Ventures, Advisory & Research, for the recognition for the fourth year in a row! And a huge congratulations to all the fellow ambassadors! I'm looking forward to collaborating with you. #ArtificialIntelligence #GenerativeAI #AgenticAI THE AI ANDY FITZE 🧠⛵ ☑️ | DALITH STEIGER-GABLINGER | Bianka Németvölgyi