Most AI demos show you the highlight reel. In this episode DK puts Claude Cowork to work on a live project management scenario: building a governance team, surfacing conflicts of interest, and drafting a change request, all from raw organisational data. No edits. The wins are real. So are the fails. And the fails are where the lesson lives. About This Episode This is a solo, screen-led demonstration episode. There is no guest. DK works through a genuine project scenario in real time using dummy data: a fictional AI build called Alliora (project P0501) inside an invented ASX-listed company, Achieve Proprietary Limited, with 4,037 staff records and a 301-project portfolio register. Watch Claude move from a document store to an actual thinking partner, and watch DK steer it with the context that separates a useful output from a confident wrong one. 💬 Key Quote “I didn't ask Claude to find these risks. I asked it to confirm who was best placed for the project. That is the difference between having a document repository and having a thinking partner.” DK, Host of Powering ProjeX 🧠 This Episode Is a Must-Listen If You... • You are curious about what AI actually does on a live project, not in a polished demo. • You suspect your organisational data is messier than anyone is willing to admit. • You sponsor or manage projects and want AI to challenge you, not just file your documents. • You are weighing up where AI belongs in governance and where humans must stay in the loop. • You want to watch a real change request, charter update and conflict of interest call play out. 🧭 What You Will Learn: 3 Key Learnings 1. AI is only as good as the data you feed it. Claude flagged the wrong sponsor, role mismatches and live conflicts of interest, not because it was clever, but because DK gave it context: the charter, the meeting minutes, the portfolio register and the employee list. The wrong sponsor was a data governance gap, not an AI failure. Most organisations have the same gap and have never looked. 2. Specificity is not optional. Vague prompts produce vague results, or worse, confident wrong ones. The discipline of being precise with AI is the same discipline that makes you a sharper project communicator, so the skill compounds in both directions. 3. A thinking partner beats a document repository. DK asked one question about team fit. Claude returned a map of risks, conflicts and structural issues she had not asked about and had not spotted. That shift, from storage to scrutiny, is the real unlock for project work. ⚙️ Most Actionable Takeaways • Audit your project and portfolio data before you point AI at it. It will find the gaps quickly and loudly. • Feed AI real context: charters, minutes, registers, role data. Context is what turns output into insight. • Use AI to pressure-test governance: sponsor clarity, conflicts of interest, decision rights, board composition. • Keep humans on the judgment calls. DK made every decision; Claude surfaced the options and the rationale. • Never upload real employee data. DK used dummy data throughout and says so plainly. Get advice from your own leadership first. 🔗 Connect & Learn More Host: Daniela Kellett | Project Director, Speaker, and Founder of Powering ProjeX LinkedIn: https://www.linkedin.com/in/daniela-kellett-813b1337/ 🧠 Takeaways • The best project ally is the one that tells you the truth, even when the truth is that your data has a gap. • AI does not fix bad data. It exposes it. • Specificity with AI is just good project communication, held to a standard. • Storage answers what you ask. A thinking partner answers what you should have asked. ⚡ Closing Thought This episode is a reminder that the quiet work of clean data, clear roles and honest governance is what lets any tool, AI included, actually help. Celebrate that work. It is the architecture of progress, even when no one is watching. Until next time, keep powering your projects from vision to victory. Subscribe to Powering ProjeX today! Listen on: YouTube: https://www.youtube.com/@poweringprojex Spotify: https://open.spotify.com/show/4ouQGB93q1OThUbNODUedu?si=23eb21e242dc4b2b Apple Podcasts: https://podcasts.apple.com/au/podcast/powering-projex-from-vision-to-victory/id1823405814 Engage with Powering ProjeX on: X: https://x.com/PoweringProjeX LinkedIn: https://www.linkedin.com/company/powering-projex/?viewAsMember=true Support the show: buymeacoffee.com/PoweringProjeX #PoweringProjeX #VisionToVictory #ProjectManagement #AIinProjects #ProjectGovernance #DataGovernance #PMO #ArtificialIntelligence #ProjectLeadership #ChangeManagement #P3M3 #AgenticAI

Most AI demos show you the highlight reel. In this episode DK puts Claude Cowork to work on a live project management scenario: building a governance team, surfacing conflicts of interest, and drafting a change request, all from raw organisational data. No edits. The wins are real. So are the fails. And the fails are where the lesson lives. About This Episode This is a solo, screen-led demonstration episode. There is no guest. DK works through a genuine project scenario in real time using dummy data: a fictional AI build called Alliora (project P0501) inside an invented ASX-listed company, Achieve Proprietary Limited, with 4,037 staff records and a 301-project portfolio register. Watch Claude move from a document store to an actual thinking partner, and watch DK steer it with the context that separates a useful output from a confident wrong one. 💬 Key Quote “I didn't ask Claude to find these risks. I asked it to confirm who was best placed for the project. That is the difference between having a document repository and having a thinking partner.” DK, Host of Powering ProjeX 🧠 This Episode Is a Must-Listen If You... • You are curious about what AI actually does on a live project, not in a polished demo. • You suspect your organisational data is messier than anyone is willing to admit. • You sponsor or manage projects and want AI to challenge you, not just file your documents. • You are weighing up where AI belongs in governance and where humans must stay in the loop. • You want to watch a real change request, charter update and conflict of interest call play out. 🧭 What You Will Learn: 3 Key Learnings 1. AI is only as good as the data you feed it. Claude flagged the wrong sponsor, role mismatches and live conflicts of interest, not because it was clever, but because DK gave it context: the charter, the meeting minutes, the portfolio register and the employee list. The wrong sponsor was a data governance gap, not an AI failure. Most organisations have the same gap and have never looked. 2. Specificity is not optional. Vague prompts produce vague results, or worse, confident wrong ones. The discipline of being precise with AI is the same discipline that makes you a sharper project communicator, so the skill compounds in both directions. 3. A thinking partner beats a document repository. DK asked one question about team fit. Claude returned a map of risks, conflicts and structural issues she had not asked about and had not spotted. That shift, from storage to scrutiny, is the real unlock for project work. ⚙️ Most Actionable Takeaways • Audit your project and portfolio data before you point AI at it. It will find the gaps quickly and loudly. • Feed AI real context: charters, minutes, registers, role data. Context is what turns output into insight. • Use AI to pressure-test governance: sponsor clarity, conflicts of interest, decision rights, board composition. • Keep humans on the judgment calls. DK made every decision; Claude surfaced the options and the rationale. • Never upload real employee data. DK used dummy data throughout and says so plainly. Get advice from your own leadership first. 🔗 Connect & Learn More Host: Daniela Kellett | Project Director, Speaker, and Founder of Powering ProjeX LinkedIn: https://www.linkedin.com/in/daniela-kellett-813b1337/ 🧠 Takeaways • The best project ally is the one that tells you the truth, even when the truth is that your data has a gap. • AI does not fix bad data. It exposes it. • Specificity with AI is just good project communication, held to a standard. • Storage answers what you ask. A thinking partner answers what you should have asked. ⚡ Closing Thought This episode is a reminder that the quiet work of clean data, clear roles and honest governance is what lets any tool, AI included, actually help. Celebrate that work. It is the architecture of progress, even when no one is watching. Until next time, keep powering your projects from vision to victory. Subscribe to Powering ProjeX today! Listen on: YouTube: https://www.youtube.com/@poweringprojex Spotify: https://open.spotify.com/show/4ouQGB93q1OThUbNODUedu?si=23eb21e242dc4b2b Apple Podcasts: https://podcasts.apple.com/au/podcast/powering-projex-from-vision-to-victory/id1823405814 Engage with Powering ProjeX on: X: https://x.com/PoweringProjeX LinkedIn: https://www.linkedin.com/company/powering-projex/?viewAsMember=true Support the show: buymeacoffee.com/PoweringProjeX #PoweringProjeX #VisionToVictory #ProjectManagement #AIinProjects #ProjectGovernance #DataGovernance #PMO #ArtificialIntelligence #ProjectLeadership #ChangeManagement #P3M3 #AgenticAI

Most AI demos show you the highlight reel. In this episode DK puts Claude Cowork to work on a live project management scenario: building a governance team, surfacing conflicts of interest, and drafting a change request, all from raw organisational data. No edits. The wins are real. So are the fails. And the fails are where the lesson lives. About This Episode This is a solo, screen-led demonstration episode. There is no guest. DK works through a genuine project scenario in real time using dummy data: a fictional AI build called Alliora (project P0501) inside an invented ASX-listed company, Achieve Proprietary Limited, with 4,037 staff records and a 301-project portfolio register. Watch Claude move from a document store to an actual thinking partner, and watch DK steer it with the context that separates a useful output from a confident wrong one. 💬 Key Quote “I didn't ask Claude to find these risks. I asked it to confirm who was best placed for the project. That is the difference between having a document repository and having a thinking partner.” DK, Host of Powering ProjeX 🧠 This Episode Is a Must-Listen If You... • You are curious about what AI actually does on a live project, not in a polished demo. • You suspect your organisational data is messier than anyone is willing to admit. • You sponsor or manage projects and want AI to challenge you, not just file your documents. • You are weighing up where AI belongs in governance and where humans must stay in the loop. • You want to watch a real change request, charter update and conflict of interest call play out. 🧭 What You Will Learn: 3 Key Learnings 1. AI is only as good as the data you feed it. Claude flagged the wrong sponsor, role mismatches and live conflicts of interest, not because it was clever, but because DK gave it context: the charter, the meeting minutes, the portfolio register and the employee list. The wrong sponsor was a data governance gap, not an AI failure. Most organisations have the same gap and have never looked. 2. Specificity is not optional. Vague prompts produce vague results, or worse, confident wrong ones. The discipline of being precise with AI is the same discipline that makes you a sharper project communicator, so the skill compounds in both directions. 3. A thinking partner beats a document repository. DK asked one question about team fit. Claude returned a map of risks, conflicts and structural issues she had not asked about and had not spotted. That shift, from storage to scrutiny, is the real unlock for project work. ⚙️ Most Actionable Takeaways • Audit your project and portfolio data before you point AI at it. It will find the gaps quickly and loudly. • Feed AI real context: charters, minutes, registers, role data. Context is what turns output into insight. • Use AI to pressure-test governance: sponsor clarity, conflicts of interest, decision rights, board composition. • Keep humans on the judgment calls. DK made every decision; Claude surfaced the options and the rationale. • Never upload real employee data. DK used dummy data throughout and says so plainly. Get advice from your own leadership first. 🔗 Connect & Learn More Host: Daniela Kellett | Project Director, Speaker, and Founder of Powering ProjeX LinkedIn: https://www.linkedin.com/in/daniela-kellett-813b1337/ 🧠 Takeaways • The best project ally is the one that tells you the truth, even when the truth is that your data has a gap. • AI does not fix bad data. It exposes it. • Specificity with AI is just good project communication, held to a standard. • Storage answers what you ask. A thinking partner answers what you should have asked. ⚡ Closing Thought This episode is a reminder that the quiet work of clean data, clear roles and honest governance is what lets any tool, AI included, actually help. Celebrate that work. It is the architecture of progress, even when no one is watching. Until next time, keep powering your projects from vision to victory. Subscribe to Powering ProjeX today! Listen on: YouTube: https://www.youtube.com/@poweringprojex Spotify: https://open.spotify.com/show/4ouQGB93q1OThUbNODUedu?si=23eb21e242dc4b2b Apple Podcasts: https://podcasts.apple.com/au/podcast/powering-projex-from-vision-to-victory/id1823405814 Engage with Powering ProjeX on: X: https://x.com/PoweringProjeX LinkedIn: https://www.linkedin.com/company/powering-projex/?viewAsMember=true Support the show: buymeacoffee.com/PoweringProjeX #PoweringProjeX #VisionToVictory #ProjectManagement #AIinProjects #ProjectGovernance #DataGovernance #PMO #ArtificialIntelligence #ProjectLeadership #ChangeManagement #P3M3 #AgenticAI

Most AI demos show you the highlight reel. In this episode DK puts Claude Cowork to work on a live project management scenario: building a governance team, surfacing conflicts of interest, and drafting a change request, all from raw organisational data. No edits. The wins are real. So are the fails. And the fails are where the lesson lives. About This Episode This is a solo, screen-led demonstration episode. There is no guest. DK works through a genuine project scenario in real time using dummy data: a fictional AI build called Alliora (project P0501) inside an invented ASX-listed company, Achieve Proprietary Limited, with 4,037 staff records and a 301-project portfolio register. Watch Claude move from a document store to an actual thinking partner, and watch DK steer it with the context that separates a useful output from a confident wrong one. 💬 Key Quote “I didn't ask Claude to find these risks. I asked it to confirm who was best placed for the project. That is the difference between having a document repository and having a thinking partner.” DK, Host of Powering ProjeX 🧠 This Episode Is a Must-Listen If You... • You are curious about what AI actually does on a live project, not in a polished demo. • You suspect your organisational data is messier than anyone is willing to admit. • You sponsor or manage projects and want AI to challenge you, not just file your documents. • You are weighing up where AI belongs in governance and where humans must stay in the loop. • You want to watch a real change request, charter update and conflict of interest call play out. 🧭 What You Will Learn: 3 Key Learnings 1. AI is only as good as the data you feed it. Claude flagged the wrong sponsor, role mismatches and live conflicts of interest, not because it was clever, but because DK gave it context: the charter, the meeting minutes, the portfolio register and the employee list. The wrong sponsor was a data governance gap, not an AI failure. Most organisations have the same gap and have never looked. 2. Specificity is not optional. Vague prompts produce vague results, or worse, confident wrong ones. The discipline of being precise with AI is the same discipline that makes you a sharper project communicator, so the skill compounds in both directions. 3. A thinking partner beats a document repository. DK asked one question about team fit. Claude returned a map of risks, conflicts and structural issues she had not asked about and had not spotted. That shift, from storage to scrutiny, is the real unlock for project work. ⚙️ Most Actionable Takeaways • Audit your project and portfolio data before you point AI at it. It will find the gaps quickly and loudly. • Feed AI real context: charters, minutes, registers, role data. Context is what turns output into insight. • Use AI to pressure-test governance: sponsor clarity, conflicts of interest, decision rights, board composition. • Keep humans on the judgment calls. DK made every decision; Claude surfaced the options and the rationale. • Never upload real employee data. DK used dummy data throughout and says so plainly. Get advice from your own leadership first. 🔗 Connect & Learn More Host: Daniela Kellett | Project Director, Speaker, and Founder of Powering ProjeX LinkedIn: https://www.linkedin.com/in/daniela-kellett-813b1337/ 🧠 Takeaways • The best project ally is the one that tells you the truth, even when the truth is that your data has a gap. • AI does not fix bad data. It exposes it. • Specificity with AI is just good project communication, held to a standard. • Storage answers what you ask. A thinking partner answers what you should have asked. ⚡ Closing Thought This episode is a reminder that the quiet work of clean data, clear roles and honest governance is what lets any tool, AI included, actually help. Celebrate that work. It is the architecture of progress, even when no one is watching. Until next time, keep powering your projects from vision to victory. Subscribe to Powering ProjeX today! Listen on: YouTube: https://www.youtube.com/@poweringprojex Spotify: https://open.spotify.com/show/4ouQGB93q1OThUbNODUedu?si=23eb21e242dc4b2b Apple Podcasts: https://podcasts.apple.com/au/podcast/powering-projex-from-vision-to-victory/id1823405814 Engage with Powering ProjeX on: X: https://x.com/PoweringProjeX LinkedIn: https://www.linkedin.com/company/powering-projex/?viewAsMember=true Support the show: buymeacoffee.com/PoweringProjeX #PoweringProjeX #VisionToVictory #ProjectManagement #AIinProjects #ProjectGovernance #DataGovernance #PMO #ArtificialIntelligence #ProjectLeadership #ChangeManagement #P3M3 #AgenticAI

Most AI demos show you the highlight reel. In this episode DK puts Claude Cowork to work on a live project management scenario: building a governance team, surfacing conflicts of interest, and drafting a change request, all from raw organisational data. No edits. The wins are real. So are the fails. And the fails are where the lesson lives. About This Episode This is a solo, screen-led demonstration episode. There is no guest. DK works through a genuine project scenario in real time using dummy data: a fictional AI build called Alliora (project P0501) inside an invented ASX-listed company, Achieve Proprietary Limited, with 4,037 staff records and a 301-project portfolio register. Watch Claude move from a document store to an actual thinking partner, and watch DK steer it with the context that separates a useful output from a confident wrong one. 💬 Key Quote “I didn't ask Claude to find these risks. I asked it to confirm who was best placed for the project. That is the difference between having a document repository and having a thinking partner.” DK, Host of Powering ProjeX 🧠 This Episode Is a Must-Listen If You... • You are curious about what AI actually does on a live project, not in a polished demo. • You suspect your organisational data is messier than anyone is willing to admit. • You sponsor or manage projects and want AI to challenge you, not just file your documents. • You are weighing up where AI belongs in governance and where humans must stay in the loop. • You want to watch a real change request, charter update and conflict of interest call play out. 🧭 What You Will Learn: 3 Key Learnings 1. AI is only as good as the data you feed it. Claude flagged the wrong sponsor, role mismatches and live conflicts of interest, not because it was clever, but because DK gave it context: the charter, the meeting minutes, the portfolio register and the employee list. The wrong sponsor was a data governance gap, not an AI failure. Most organisations have the same gap and have never looked. 2. Specificity is not optional. Vague prompts produce vague results, or worse, confident wrong ones. The discipline of being precise with AI is the same discipline that makes you a sharper project communicator, so the skill compounds in both directions. 3. A thinking partner beats a document repository. DK asked one question about team fit. Claude returned a map of risks, conflicts and structural issues she had not asked about and had not spotted. That shift, from storage to scrutiny, is the real unlock for project work. ⚙️ Most Actionable Takeaways • Audit your project and portfolio data before you point AI at it. It will find the gaps quickly and loudly. • Feed AI real context: charters, minutes, registers, role data. Context is what turns output into insight. • Use AI to pressure-test governance: sponsor clarity, conflicts of interest, decision rights, board composition. • Keep humans on the judgment calls. DK made every decision; Claude surfaced the options and the rationale. • Never upload real employee data. DK used dummy data throughout and says so plainly. Get advice from your own leadership first. 🔗 Connect & Learn More Host: Daniela Kellett | Project Director, Speaker, and Founder of Powering ProjeX LinkedIn: https://www.linkedin.com/in/daniela-kellett-813b1337/ 🧠 Takeaways • The best project ally is the one that tells you the truth, even when the truth is that your data has a gap. • AI does not fix bad data. It exposes it. • Specificity with AI is just good project communication, held to a standard. • Storage answers what you ask. A thinking partner answers what you should have asked. ⚡ Closing Thought This episode is a reminder that the quiet work of clean data, clear roles and honest governance is what lets any tool, AI included, actually help. Celebrate that work. It is the architecture of progress, even when no one is watching. Until next time, keep powering your projects from vision to victory. Subscribe to Powering ProjeX today! Listen on: YouTube: https://www.youtube.com/@poweringprojex Spotify: https://open.spotify.com/show/4ouQGB93q1OThUbNODUedu?si=23eb21e242dc4b2b Apple Podcasts: https://podcasts.apple.com/au/podcast/powering-projex-from-vision-to-victory/id1823405814 Engage with Powering ProjeX on: X: https://x.com/PoweringProjeX LinkedIn: https://www.linkedin.com/company/powering-projex/?viewAsMember=true Support the show: buymeacoffee.com/PoweringProjeX #PoweringProjeX #VisionToVictory #ProjectManagement #AIinProjects #ProjectGovernance #DataGovernance #PMO #ArtificialIntelligence #ProjectLeadership #ChangeManagement #P3M3 #AgenticAI

Most project managers know AI can help. But knowing that and actually having it set up, loaded with context, and ready to execute meaningful work are two very different things. In this episode, DK walks you through exactly how she has configured Claude CoWork from scratch: the desktop setup, projects, skills files, and voice prompting, then runs a live trial on a real (dummy) portfolio strategy to show what AI-assisted project decision-making actually looks like in practice. After running a live portfolio strategy trial analysing 301 projects against a $20 million CapEx constraint, DK is clear: AI is outstanding at synthesising large amounts of structured data and giving you a starting position. But real-time reprioritising, reading the room, and navigating political complexity still requires a human. The goal is partnership, not replacement. 💬 Key Quote "I know that it cannot do everything that I do automagically. It just can't. But there is definitely space for AI, and definitely still space for humans. It is just about how we synthesise all of that and create this perfect partnership." — DK (Daniela Kellett) 🧠 This Episode Is a Must-Listen If You... • Have tried AI tools but felt like you were starting from zero every time and never really got traction. • Are responsible for portfolio decisions and want to see how AI can help you think through stop, start, and continue choices. • Work in project delivery and want to understand the practical difference between AI as a chat tool and AI as a genuine project ally. 🧭 What You Will Learn: 3 Key Learnings Learning 1: Context is everything. If you do not set it up, AI cannot help you. Claude is only as useful as the context you give it. DK walks through exactly how she structures her Claude projects with detailed instructions, attached files, and markdown skills files. Learning 2: Skills files are your secret weapon for repeatability. Creating markdown (.md) skills files allows you to embed your methodology once and reuse it across every engagement. Learning 3: AI cannot replace in-the-moment human judgment, but it can dramatically reduce manual work. ⚙️ Most Actionable Takeaways • Create at least one Claude project for your main work area. Write clear instructions: who you are, what you do, what your standards are, and what you need from it. Attach relevant documents so it has real context. • Build markdown skills files for your core delivery disciplines. • Set up Claude on your phone and use voice prompting. • If designing an automated project workflow: N8N for scheduling and triggers, Supabase for structured project data, and Claude for the intelligence and analysis layer. • Run a stop, start, continue exercise on your own portfolio. 🔗 Connect and Learn More Host: Daniela Kellett (DK) | Project Director, Speaker, Founder of Powering ProjeX and Alliora.AI LinkedIn: https://www.linkedin.com/in/daniela-kellett-813b1337/ Alliora.AI: https://alliora.ai 🧠 Takeaways • Your AI is only as smart as the context you give it. Set it up properly once and it pays dividends forever. • AI cannot replace the in-the-moment thinking of a great project manager, but it can eliminate hours of manual work every week. • The partnership between human judgment and AI capability is not a future state. It is available right now, if you build it. ⚡ Closing Thought This episode is a blueprint, not for a perfect AI system, but for the mindset shift that every project professional needs to make: AI is not something that happens to you, it is something you build with intention. DK has done the hard work of figuring out what works, and she is sharing it so you do not have to start from zero. The project professionals who invest in giving their AI the context and structure it needs to be genuinely useful are the ones who will be delivering faster, smarter, and with far less waste in the years ahead. Until next time, keep powering your projects from vision to victory. Subscribe to Powering ProjeX Today! If this episode gave you one idea you can use this week, share it with a project manager who needs to hear it. Listen on: YouTube: https://www.youtube.com/@poweringprojex Spotify: https://open.spotify.com/show/4ouQGB93q1OThUbNODUedu?si=23eb21e242dc4b2b Apple Podcasts: https://podcasts.apple.com/au/podcast/powering-projex-from-vision-to-victory/id1823405814 Engage with Powering ProjeX on: X: https://x.com/PoweringProjeX LinkedIn: https://www.linkedin.com/company/powering-projex/?viewAsMember=true #PoweringProjeX #VisionToVictory #ProjectManagement #AIinProjects #ClaudeAI #ProjectDelivery #PortfolioManagement #AgenticAI #PMO #ProjectManager #AItools #DigitalTransformation #ProjectLeadership

Most project managers know AI can help. But knowing that and actually having it set up, loaded with context, and ready to execute meaningful work are two very different things. In this episode, DK walks you through exactly how she has configured Claude CoWork from scratch: the desktop setup, projects, skills files, and voice prompting, then runs a live trial on a real (dummy) portfolio strategy to show what AI-assisted project decision-making actually looks like in practice. After running a live portfolio strategy trial analysing 301 projects against a $20 million CapEx constraint, DK is clear: AI is outstanding at synthesising large amounts of structured data and giving you a starting position. But real-time reprioritising, reading the room, and navigating political complexity still requires a human. The goal is partnership, not replacement. 💬 Key Quote "I know that it cannot do everything that I do automagically. It just can't. But there is definitely space for AI, and definitely still space for humans. It is just about how we synthesise all of that and create this perfect partnership." — DK (Daniela Kellett) 🧠 This Episode Is a Must-Listen If You... • Have tried AI tools but felt like you were starting from zero every time and never really got traction. • Are responsible for portfolio decisions and want to see how AI can help you think through stop, start, and continue choices. • Work in project delivery and want to understand the practical difference between AI as a chat tool and AI as a genuine project ally. 🧭 What You Will Learn: 3 Key Learnings Learning 1: Context is everything. If you do not set it up, AI cannot help you. Claude is only as useful as the context you give it. DK walks through exactly how she structures her Claude projects with detailed instructions, attached files, and markdown skills files. Learning 2: Skills files are your secret weapon for repeatability. Creating markdown (.md) skills files allows you to embed your methodology once and reuse it across every engagement. Learning 3: AI cannot replace in-the-moment human judgment, but it can dramatically reduce manual work. ⚙️ Most Actionable Takeaways • Create at least one Claude project for your main work area. Write clear instructions: who you are, what you do, what your standards are, and what you need from it. Attach relevant documents so it has real context. • Build markdown skills files for your core delivery disciplines. • Set up Claude on your phone and use voice prompting. • If designing an automated project workflow: N8N for scheduling and triggers, Supabase for structured project data, and Claude for the intelligence and analysis layer. • Run a stop, start, continue exercise on your own portfolio. 🔗 Connect and Learn More Host: Daniela Kellett (DK) | Project Director, Speaker, Founder of Powering ProjeX and Alliora.AI LinkedIn: https://www.linkedin.com/in/daniela-kellett-813b1337/ Alliora.AI: https://alliora.ai 🧠 Takeaways • Your AI is only as smart as the context you give it. Set it up properly once and it pays dividends forever. • AI cannot replace the in-the-moment thinking of a great project manager, but it can eliminate hours of manual work every week. • The partnership between human judgment and AI capability is not a future state. It is available right now, if you build it. ⚡ Closing Thought This episode is a blueprint, not for a perfect AI system, but for the mindset shift that every project professional needs to make: AI is not something that happens to you, it is something you build with intention. DK has done the hard work of figuring out what works, and she is sharing it so you do not have to start from zero. The project professionals who invest in giving their AI the context and structure it needs to be genuinely useful are the ones who will be delivering faster, smarter, and with far less waste in the years ahead. Until next time, keep powering your projects from vision to victory. Subscribe to Powering ProjeX Today! If this episode gave you one idea you can use this week, share it with a project manager who needs to hear it. Listen on: YouTube: https://www.youtube.com/@poweringprojex Spotify: https://open.spotify.com/show/4ouQGB93q1OThUbNODUedu?si=23eb21e242dc4b2b Apple Podcasts: https://podcasts.apple.com/au/podcast/powering-projex-from-vision-to-victory/id1823405814 Engage with Powering ProjeX on: X: https://x.com/PoweringProjeX LinkedIn: https://www.linkedin.com/company/powering-projex/?viewAsMember=true #PoweringProjeX #VisionToVictory #ProjectManagement #AIinProjects #ClaudeAI #ProjectDelivery #PortfolioManagement #AgenticAI #PMO #ProjectManager #AItools #DigitalTransformation #ProjectLeadership

Most project managers know AI can help. But knowing that and actually having it set up, loaded with context, and ready to execute meaningful work are two very different things. In this episode, DK walks you through exactly how she has configured Claude CoWork from scratch: the desktop setup, projects, skills files, and voice prompting, then runs a live trial on a real (dummy) portfolio strategy to show what AI-assisted project decision-making actually looks like in practice. After running a live portfolio strategy trial analysing 301 projects against a $20 million CapEx constraint, DK is clear: AI is outstanding at synthesising large amounts of structured data and giving you a starting position. But real-time reprioritising, reading the room, and navigating political complexity still requires a human. The goal is partnership, not replacement. 💬 Key Quote "I know that it cannot do everything that I do automagically. It just can't. But there is definitely space for AI, and definitely still space for humans. It is just about how we synthesise all of that and create this perfect partnership." — DK (Daniela Kellett) 🧠 This Episode Is a Must-Listen If You... • Have tried AI tools but felt like you were starting from zero every time and never really got traction. • Are responsible for portfolio decisions and want to see how AI can help you think through stop, start, and continue choices. • Work in project delivery and want to understand the practical difference between AI as a chat tool and AI as a genuine project ally. 🧭 What You Will Learn: 3 Key Learnings Learning 1: Context is everything. If you do not set it up, AI cannot help you. Claude is only as useful as the context you give it. DK walks through exactly how she structures her Claude projects with detailed instructions, attached files, and markdown skills files. Learning 2: Skills files are your secret weapon for repeatability. Creating markdown (.md) skills files allows you to embed your methodology once and reuse it across every engagement. Learning 3: AI cannot replace in-the-moment human judgment, but it can dramatically reduce manual work. ⚙️ Most Actionable Takeaways • Create at least one Claude project for your main work area. Write clear instructions: who you are, what you do, what your standards are, and what you need from it. Attach relevant documents so it has real context. • Build markdown skills files for your core delivery disciplines. • Set up Claude on your phone and use voice prompting. • If designing an automated project workflow: N8N for scheduling and triggers, Supabase for structured project data, and Claude for the intelligence and analysis layer. • Run a stop, start, continue exercise on your own portfolio. 🔗 Connect and Learn More Host: Daniela Kellett (DK) | Project Director, Speaker, Founder of Powering ProjeX and Alliora.AI LinkedIn: https://www.linkedin.com/in/daniela-kellett-813b1337/ Alliora.AI: https://alliora.ai 🧠 Takeaways • Your AI is only as smart as the context you give it. Set it up properly once and it pays dividends forever. • AI cannot replace the in-the-moment thinking of a great project manager, but it can eliminate hours of manual work every week. • The partnership between human judgment and AI capability is not a future state. It is available right now, if you build it. ⚡ Closing Thought This episode is a blueprint, not for a perfect AI system, but for the mindset shift that every project professional needs to make: AI is not something that happens to you, it is something you build with intention. DK has done the hard work of figuring out what works, and she is sharing it so you do not have to start from zero. The project professionals who invest in giving their AI the context and structure it needs to be genuinely useful are the ones who will be delivering faster, smarter, and with far less waste in the years ahead. Until next time, keep powering your projects from vision to victory. Subscribe to Powering ProjeX Today! If this episode gave you one idea you can use this week, share it with a project manager who needs to hear it. Listen on: YouTube: https://www.youtube.com/@poweringprojex Spotify: https://open.spotify.com/show/4ouQGB93q1OThUbNODUedu?si=23eb21e242dc4b2b Apple Podcasts: https://podcasts.apple.com/au/podcast/powering-projex-from-vision-to-victory/id1823405814 Engage with Powering ProjeX on: X: https://x.com/PoweringProjeX LinkedIn: https://www.linkedin.com/company/powering-projex/?viewAsMember=true #PoweringProjeX #VisionToVictory #ProjectManagement #AIinProjects #ClaudeAI #ProjectDelivery #PortfolioManagement #AgenticAI #PMO #ProjectManager #AItools #DigitalTransformation #ProjectLeadership