The Sovereign Career Hub

Human Led, AI Forward.
  • 0.0 (0)
  • Careers
  • Updated Weekly

The Sovereign Career Hub offers practical AI workflows, video demos, prompts and lived experience that build your judgement about where AI belongs, where it is risky, and where it is not needed at all. Hosted by Carolyn Shepherd, champion of human judgement in the age of AI, creator of Jobscaping and one of HR Magazine's Most Influential Thinkers 2025. carolynshepherd.substack.com

  1. 5d ago

    Named: 68 Steps to hand over to Copilot Analyst

    Show Notes: Ep 14 Named: 68 Steps to Hand to Copilot AnalystWebsite: [merrillconsultants.com](https://www.merrillconsultants.com/)LinkedIn: [linkedin.com/in/merrillvalerie](https://www.linkedin.com/in/merrillvalerie/)Mentioned in this episodeJobscaping®Carolyn's original framework for mapping your work into three zones, green to delegate, amber to collaborate, red to protect, so you lead AI rather than hand your judgement away. The full course is coming soon. In the meantime, read Carolyn's Jobscaping® article on LinkedIn.[AI-Ready Jobscaping: Why Our Careers Need More Than a One-and-Done Mindset](https://www.linkedin.com/pulse/ai-ready-jobscaping-why-our-careers-need-more-than-mindset-shepherd-nadme/)Microsoft Copilot AnalystThe second of Microsoft's two deep reasoning agents, alongside Researcher. If Researcher is your AI consultant, Analyst is your data scientist. It runs calculations, identifies trends and anomalies, builds visualisations, generates forecasts, produces analytical reports and can use Python for advanced work. It shows its working as it goes, and it sits inside the apps you already use. No coding required.microsoft.com/en-gb/microsoft-copilot Your monthly research budgetResearcher and Analyst share 25 queries between them, per calendar month. Spend them on the questions that are actually worth the depth.The companion screens guideA step by step walkthrough with screenshots of the Analyst prompt, the payroll insights and the finished charts, so you can follow along even if you're only listening. It covers Researcher too.Researcher and Analyst companion guideThe Sovereign Career Hub Everything on the Hub is free, so if you've missed an episode you can catch up any time. Subscribe and you'll know when the next one lands.Substack: https://substack.com/@thesovereigncareerhubCarolyn on LinkedIn: https://www.linkedin.com/in/carolynjshepherd/ Emmeline AI: emmeline.ai](https://www.emmeline.ai The Problem Analyst Solves You’ve done the analysis. You’ve sent the spreadsheet. And nothing happens. That isn’t a numbers problem. It’s a translation problem. The figures were right, they just never became something anybody else could act on. Get that part working and you stop being the person who sends the spreadsheets. You become the person in the room who gets what they asked for. The Basic Agent This week Valerie Merrill introduces Analyst, the second of Copilot’s two deep reasoning agents. If last week’s Researcher agent is your AI consultant, then this week’s Analyst agent is your data scientist. You upload a file, you ask a question, and it works through it logically. In the episode it took a payroll file covering 287 employees and came back after 68 reasoning steps with headcount, gross pay, the real cost drivers and a set of charts. Sixty eight steps. Many of which none of us would have thought to take. The Agent’s True Value The part worth staying for comes next. We asked it a leader’s question rather than a data question: ‘If I were head of payroll, what should I recommend to the board? It refused the obvious answer. Rather than hunting for cost savings, it said the payroll operation looked well controlled and suggested framing the conversation around workforce planning, governance and risk instead. Then we asked it to write that up as a narrative for a conversation with the finance director, in the FD’s language. The result: Context, analysis, communications. Three different skills. Not so long ago that was three different people. The Human Element One caution, and it matters more than anything else in the episode. Analyst is fluent, but fluent is not the same as correct. It’s predictive, it’s probabilistic, and it can be confidently wrong. You need enough domain experience to interrogate what it hands you, which is exactly why it can make you look good and can’t replace you. The analysis is its job. The judgement is still yours. Valerie Merrill is a training and development specialist, director of Merrill Consultants and a Fellow of the Learning and Performance Institute. She’s also someone I’m glad to call a colleague and a friend. Your Challenge Take one set of numbers you need somebody else to care about, and let Analyst help you build the case. Thanks for reading! Subscribe for free to receive new posts and support my work. Everything on the Hub is free, so if you’ve missed an episode you can catch up any time. Subscribing means you’ll know when I publish these posts, and when the articles drop plus you can eaves drop on, or contribute to, the chat screens. Full show notes and the companion screens guide are below. The guide has screenshots of the key steps, so you can follow along even if you’re only listening: Researcher and Analyst companion guide Next week Valerie shows you how to build your own agent in Copilot. This is a skill we will all need very soon. Mainly so we can delegate the work nothing off the shelf will touch.So join us next week for the tour and I’ll add some Jobscaping tips for spotting the Zone one (Jobscaping) green work that’s worth delegating. Thanks for reading, now click to listen… And until next week, stay sovereign. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit carolynshepherd.substack.com

  2. Jul 29

    Why Pay AI Tax? Copilot Researcher Agent May Be The Answer

    Valerie Merrill Valerie Merrill is a Microsoft and people skills specialist, Fellow of the Learning and Performance Institute, award-winning L&D podcaster and one of the UK's most experienced Copilot trainers. Website: https://www.merrillconsultants.com/ LinkedIn: https://www.linkedin.com/in/merrillvalerie/ Jobscaping® Carolyn Shepherd's framework for helping people consciously redesign their work in the age of AI. Using three simple zones, Green (delegate), Amber (collaborate) and Red (protect), Jobscaping® helps professionals decide where AI can create value and where human judgement remains essential. Read more: https://www.linkedin.com/pulse/ai-ready-jobscaping-why-our-careers-need-more-than-mindset-shepherd-nadme/ Microsoft Copilot Researcher Researcher is one of Microsoft's deep reasoning agents. It investigates complex topics, draws on multiple sources and produces detailed, referenced reports that can be exported into formats such as Word, PowerPoint and PDF. https://drive.google.com/file/d/1CV-4r_1-SjWr3CFUtGD2KzAUyNp0_2s5/view?usp=sharing Companion Guide Follow along with the screenshots and walkthrough featured in the episode: https://drive.google.com/file/d/1CV-4r_1-SjWr3CFUtGD2KzAUyNp0_2s5/view?usp=sharing The Sovereign Career Hub https://substack.com/@thesovereigncareerhub https://www.linkedin.com/in/carolynjshepherd/ Emmeline AI Transcript Hello and welcome to this episode which promises to move your AI efficiency rating up a couple of notches. One of AI’s biggest promises is saving us time. Yet many people find themselves paying so-called AI Tax: rewriting prompts, adding context, chasing sources and asking endless follow-up questions, only to end up doing much of the research themselves. If you’ve ever asked Copilot a question and felt that the answer gave you the gist when you needed genuine depth, this episode is for you. This week, learning and development specialist Valerie Merrill demonstrates Researcher, one of Microsoft’s deep reasoning agents. She has kindly shared some screenshots here if you need them to accompany this video. Unlike a standard Copilot chat, Researcher is designed to investigate complex questions, review multiple sources and produce a structured, referenced report that can support real decision-making and the good news is - it’s easy to use and probably included with your current license. Used well, Researcher can save hours of searching, reading and prompt refinement. But here’s the important part: The goal isn’t simply to work faster. It’s to spend less time gathering information and more time applying your judgement. This Week’s Jobscaping® Challenge Take one research task from your Green Zone list. It might be: * Researching a market or industry trend * Comparing suppliers or solutions * Preparing a recommendation for stakeholders * Investigating a new technology or business opportunity Before opening Researcher, create a clear brief. Think about: * Your objective * Your audience * The context * The sources you would like considered If you’re struggling to structure your request, revisit last week’s episode on Prompt Coach first and use it for a quick win. Arguably, a good brief reduces AI Tax by 50% and a great brief eliminates it. More on this in the longer article to accompany this podcast. How Researcher Works * Open MS Copilot and select Researcher * Provide a clear brief with your objective, audience, context and relevant sources * Answer any follow-up questions it asks * Review the report, check the citations and read key source material for yourself Researcher can gather information. Judgement remains your responsibility. Why This Matters Researcher is a great example of Jobscaping® in action. Rather than repeatedly prompting a general-purpose chatbot and paying so-called AI Tax, you’re choosing a specialist tool designed for deeper investigation from the outset. * Less time searching. * Less time re-prompting. * More time making better decisions. A Word of Caution A professional-looking report is not necessarily a correct report. Researcher provides citations for a reason. Follow them, especially when the findings will influence important decisions, recommendations or actions. Thanks for reading! This post is public so feel free to share it. Coming Up Next Week Next week Valerie introduces Analyst, Researcher’s data-focused counterpart. If Researcher helps you investigate ideas, Analyst helps you explore spreadsheets, datasets and numerical information with the same depth of reasoning. Together, they represent a significant step beyond traditional chatbot-style AI. Best wishes Carolyn Shepherd This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit carolynshepherd.substack.com

  3. Jul 22

    Ever wondered if an AI agent could help you prompt better?

    Shownotes: Valerie Merrill, FLPI.Learning and Development SpecialistDirector, Merrill Consultants Website: merrillconsultants.comLinkedIn: linkedin.com/in/merrillvalerie The Sovereign Career Hub Substack: substack.com/@thesovereigncareerhub Carolyn on LinkedIn: linkedin.com/in/carolynjshepherd Emmeline AI: emmeline.ai Jobscaping® Carolyn’s original framework for mapping your work into three zones, green to delegate, amber to collaborate, red to protect, so you lead AI rather than hand your judgement away. The full course is coming soon. In the meantime, read Carolyn’s Jobscaping® article on LinkedIn. First episode in this mini-series Setting the scene for Jobscaping - Human Led: AI Forward https://substack.com/@thesovereigncareerhub/note/p-207200973?r=3osjv8&utm_source=notes-share-action&utm_medium=web Companion Article Learn How to Map Your Workflows https://substack.com/@thesovereigncareerhub/note/c-298177681?r=3osjv8&utm_source=notes-share-action&utm_medium=web AI-Ready Jobscaping: Why Our Careers Need More Than a One-and-Done Mindset https://www.linkedin.com/pulse/ai-ready-jobscaping-why-our-careers-need-more-than-mindset-shepherd-nadme/ ------------- The practical follow on to last week. You marked your green zone work. This week I guide you to start breaking down your workflows into tasks while Learning and Development specialist, Valerie Merrill shows you how to hand the repeatable parts to Microsoft’s ready made AI agent, Prompt Coach. Noting that a better prompt gets you a better answer the first time. Watch the video, try it out, and you decide. This week’s step Take one automatable Z1 job from your list and pull it apart. A job is almost never one thing. Your monthly report feels like a single task, but look closely and it is ten or eleven small steps: gathering the numbers, cleaning them up, checking against last month, dropping it into the template. Those repeatable steps are exactly what an agent is built to take off your hands. There is also a companion article on the Hub that breaks this workflow down further (and I signpost you to it at the relevant moment from this podcast so no need to look for it yet.) How Prompt Coach works, in four steps You are guided through each step and given takeaway tips. In a nutshell, you will be shown these steps first, open Copilot and find the AI agents. Second, open Prompt Coach and tell it the job you are trying to do. Third, let it sharpen your prompt so your intention is right. Fourth, copy that prompt and run it in Copilot. Then check the result in your own voice before you use it. The lesson is not the one-pager. It is that a better prompt gives you a better answer every time. Using AI responsibly A report is only as good as the source behind it, and only as trustworthy as your own check of it. Never send out something you have not read yourself, especially to a client or a decision-maker. Let the AI draft but remember you will be held accountable. There is more on developing judgement as part of the Jobscaping® course being published in Autumn. Next Week Next week Valerie goes a step further and puts AI to work as a researcher, showing you how Copilot can gather and make sense of information for you in a more powerful way than using the usual Chat function. Valerie is a Microsoft and people skills specialist and a Fellow of the Learning Professionals Institute and an award-winning L&D podcaster. More details in the Shownotes. See you next week when we dive into Copilot’s Researcher agent. It’s one of my favourites and I think you’ll find it useful. Stay Sovereign. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit carolynshepherd.substack.com

  4. Jul 15

    Are you worrying about using AI?

    The First Part of the New Series If work is arriving faster than you can clear it, and you have been handing work to AI because everyone says you should, but without a real plan. If you have a creeping feeling that soon nothing will quite sound like you, and that you could be held accountable for work you didn’t do. It seems unfair, and yet the alternative, falling behind in the AI skills race, feels worse.If this is how you feel, you are not alone and this new series was written for you. This is the first episode, and here is how it can help.As AI continues moving into the work you do, this series helps you lead it rather than be led by it, and stay in charge of your own career while everyone around you hands bits of theirs away without thinking it through. That is where Jobscaping® comes in. Right now, Jobscaping® is about mapping your work into three zones. You take what you actually do, lay it out as a line of tasks, and name each one. Green is the work you are glad to hand to AI, once you have seen for yourself that it is safe to let go. Amber is the work you keep your hands on, with AI helping. Red is the work that is authentically yours. The moment it is named, you know you would never give it away.Some things look like ‘busy’ work but are not, because they carry your judgement, your relationships, your voice. You feel the difference the instant you hear it. That instinct is the whole point. Your first step, before next week, needs no new tools. Write down everything you do in a typical week. Don’t change anything. Don’t judge it. Just get it out of your head and onto the page. That list is your raw material, and everything in this series works off it.Together we can identify the obvious green zone work that you can hand to AI. It is likely to be boring, busy work, but it will be up to you what you hand to AI and what you keep for you. The Second Part of the New Series But once you have decided to hand some work to AI, the next challenge is how to hand it over. So for the next few episodes I’m joined by Valerie Merrill, a Microsoft expert who will take you through this territory on screen, step by step, in accessible 15 minute demos here on the Sovereign Career Hub.So starting next week Valerie will introduce you to Microsoft’s ready-made agents. There is no tech jargon to worry about, and this isn’t about handing everything to AI either. It’s about handing some of the tasks you do to AI, safely and at your pace.Listen in for details and… breathe! Take care for now and stay Sovereign. Thanks for reading! Subscribe for free to receive new posts and support my work. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit carolynshepherd.substack.com

  5. Jul 1

    Using AI: What's In It For You?

    Before you touch the technology, you have to answer one question.Why would I use this?Not as a policy. Not as a business case. As a person.That is the question Erica Farmer, this week’s guest, puts at the centre of every AI transformation conversation she has. And it is the question most organisations skip entirely.Erica is founder of QuantumRise, a keynote speaker and futurist on people-first AI, and the author of the newly published AI for People Professionals. She has spent her career working with HR and L&D leaders inside large, complex organisations, and she is clear on what the evidence tells us: AI adoption fails when it treats people as users rather than as people. Her phrase for the fix is the AI dividend.Not ROI. Not productivity gain. The personal, specific, meaningful thing that makes someone want to show up differently because of AI. For Erica, who has fibromyalgia, it is the ability to run two businesses without burning out. For someone else it could be the headspace to do the work that actually matters. Get that right and engagement follows. Miss it and you just have an email strategy with “AI” in the subject line. We also get into the question of who drives transformation in an organisation. Erica's answer is not top-down alone and not grassroots alone. It is both at once. The top layer and the frontline both need to become AI literate first. Then the management middle layer is what makes it real, supercharging change through coaching, enablement and support. Not as a bolt-on. Not as an email recipient. As the people who activate everything the organisation has invested in above and below them.**Have you found your AI dividend yet? And have you helped the people around you find theirs?**The article follows later in the week, the resource is obviously, Erica’s book - here’s the link if you haven’t purchased it already AI for People Professionals . A solo audio walk through would be overkill this week so it just remains for me to invite you to join me next week for some AI reflection when I will be asking questions like: ‘what have we learned from the guests who have shared their insights from the real world?’ ‘Have they helped us tackle AI more boldly?’ and ‘What have they said that made us think harder about our relationship with AI? Until then, thank you for listening, watching and engaging - you make it all worthwhile. Take care and stay Sovereign. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit carolynshepherd.substack.com

  6. Jun 24

    First the Data, Then the AI Agent

    The agent is not the starting point.The data is.That is what every business rushing toward AI implementation is missing. And it is costing them more than they realise.Jonny Winn is AI Strategy Director at The One Point, a Yorkshire-based managed service provider working with SMBs, enterprise and public sector organisations. He came to this role through fifteen years in creative production, a business degree, and a decision at forty to go back and do an MSc in AI and data science at Hull University, where he has just received a distinction. He is new to the practice of this work. That is exactly why I wanted to talk to him.His view from inside real organisations is clear. Business leaders want AI. They want agents, automation, efficiency. And those things are genuinely available, and genuinely transformative, when the foundations are in place. But most organisations are trying to build the agent before they have sorted the data. And without clean, structured, accessible data, the agent has nothing solid to work from. “Without the data foundations, it’s just going to become expensive experimentation.” The good news is that agentic AI, built on a proper knowledge base, is within reach for organisations of every size, without a technical team, using tools they already pay for. Jonny has built a demonstration to prove it. And he has produced a video walkthrough as this week’s resource, so you can see exactly how to do it yourself. I’ll link it for download (this week’s resource) in the article on Monday as usual. Is your organisation queuing for the agent while the data foundations are still waiting?The companion article and Jobscaping® reflection follow later in the week, and my solo audio walks through Jonny’s demonstration video with you.Take care for now and stay Sovereign! This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit carolynshepherd.substack.com

  7. Jun 17

    L&D's Four Enablers with Jo Cook. Ep 08

    Show Notes: Ep 08 – L&D's Four Enablers with Jo Cook Jo Cook Editor, Training Journal Director, Lightbulb Moment Training Journal: trainingjournal.com Jo's author page: trainingjournal.com/author/jocook Lightbulb Moment: lightbulbmoment.info LinkedIn: linkedin.com/in/jocooklightbulb Mentioned in this episode TJ L&D Influence Report 2026 Introduces the Readiness Enablers Index and highlights practical conditions including stakeholder access, data, experimentation and support. Free to download. Includes a link to the 2026 survey. Download here TJ Readiness Survey (add your data): linked from the Training Journal homepage at trainingjournal.com Nigel Harrison, How to be a True Business Partner by Performance Consulting: amazon.co.uk Cathy Moore, Map It: The hands-on guide to strategic training design: amazon.com The Sovereign Career Hub Substack: substack.com/@thesovereigncareerhub Carolyn on LinkedIn: linkedin.com/in/carolynjshepherd Emmeline AI: emmeline.ai ------------------------------------------------------ Jo is the editor of Training Journal, the profession’s longest-running publication, and the author of the TJ L&D Influence Report 2026. She built it from the voices of practitioners gathered at the TJ 60th Anniversary Conference, at which I delivered the keynote. It is one of the most useful pieces of research the profession has produced this year. Thanks for reading! but don’t keep it to yourself. Do you know someone who would enjoy reading this post? Jo’s finding is not that L&D lacks knowledge. The sector has known what good looks like for a long time. Her finding is that most L&D professionals are trying to act without the conditions in place to act. She calls them the four enablers. And her argument, backed by the research and by a candid story from her own career, is that if any one of them is missing, the whole thing stalls. This conversation follows naturally from last week’s episode with Andrew Jacobs, and it looks at the same challenge from a different angle. Together they are worth sitting with. Which of the four enablers do you currently have? Which will you start work on this week? The companion article, the resource and the solo audio follow later in the week as always. Listen in, enjoy and take care for now. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit carolynshepherd.substack.com

  8. Jun 10

    Principles Over Strategy with Andrew Jacobs

    Thank you to my guest Andrew Jacobs Andrew is a learning strategist and organisational impact advisor, and founder of Llarn Learning. He is a Fellow of the Learning Performance Institute (FLPI). His work focuses on shifting organisations away from training delivery and towards capability, performance and measurable impact. His experience spans private and public sector organisations, from financial services and retail through to local and central government. He is known for a direct, principled critique of how learning functions operate, and for building the conditions where learning happens through work, not around it. He speaks internationally on learning, capability and organisational change. He produces Women Talking About Learning, a podcast listened to in over 46 countries, and writes daily at lostanddesperate.com. TranscriptYour strategy is already out of date. Not because you wrote it badly. Because the pace of change has outrun it. That is the challenge Andrew Jacobs brings to this conversation. And his answer is not to write a better one. Andrew is a well respected voice in the L&D sector, a learning strategist and organisational impact advisor, founder of Llarn Learning and chartered fellow of the CIPD. He has spent his career helping organisations move beyond training delivery and into genuine capability and doesn’t mince his words when critiquing the status quo: “…we've become shopkeepers." Someone in the business asks for a course. L&D supplies it. The doors close. Nobody checks whether anything actually changed. Andrew argues the function needs to become something different and his advice is super clear - we need to become engineers. Listen to this week’s guest interview with Andrew to discover how to transform your L&D function and to convert your team into ‘Engineers’. ie. practitioners who diagnose before they prescribe, who measure impact at 10 days, 10 weeks and 10 months, and who hold the organisation accountable for the outcomes it said it wanted. And in a world where AI is reshaping work faster than any strategy was built to track, Andrew has something more durable to offer than a strategy document. He shows us how to lean into ‘Principles’ and this surprisingly simple approach could hold the key to your sovereign career. Click play to hear the full story. PS. …don’t forget to check out the ‘three little words’ that will change every conversation you have with colleagues asking for courses. As usual, the companion article and a Jobscaping™ reflection resource follow later in the 7-day cycle, and my solo audio closes out the week. Thanks for reading! Please share this free post to support my work - This is how the Hub grows. Take care for now and Stay Sovereign! This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit carolynshepherd.substack.com

See All (20)

About

The Sovereign Career Hub offers practical AI workflows, video demos, prompts and lived experience that build your judgement about where AI belongs, where it is risky, and where it is not needed at all. Hosted by Carolyn Shepherd, champion of human judgement in the age of AI, creator of Jobscaping and one of HR Magazine's Most Influential Thinkers 2025. carolynshepherd.substack.com

Information

  • Creator
    Human Led, AI Forward.
  • Years Active
    2026
  • Episodes
    20
  • Rating
    Clean
  • Copyright
    © Carolyn Shepherd
  • Show Website
    The Sovereign Career Hub