Nap Stack

Monica Ly

A five-minute podcast for Australian business leaders navigating AI and data without a technical background. Each episode is one idea you can take into your next meeting: a question worth asking, a claim worth challenging, or a decision worth slowing down. Hosted by Monica Ly, co-founder of EdgeRed (part of The Omnia Collective) — eight years of hands on delivery experience, recorded between baby feeds and naps.

Episodes

  1. 1 day ago

    EP10: AI won't break your tech. It'll break your operating model | AI ownership & operating models

    Most leaders still treat AI as a technology problem. The ones getting stuck are the ones who never moved past that framing. Anthony Saba, partner at ADAPT and someone who speaks to hundreds of Australian executives every year, joins Monica to explain why the thing AI is most likely to break isn't the technology, it's the operating model you built the business around. His take on the AI ownership question everyone's debating: it has to be CEO-led, then three executives need to be aligned like the three keys on a nuclear submarine. The CIO or CTO owns the tech foundations. The CFO, because AI challenges how work gets funded. The COO, to orchestrate the workflow redesign. His advice on where to start: don't treat it as an AI use case. Pick one meaningful end-to-end workflow and redesign it from first principles, with intelligence built in from the start, not bolted on at the edges. You learn far more from that than from proving something is technically possible. Five minutes. One idea. Something to bring to your next leadership conversation. Have a question or topic for Monica? Visit edgered.com.au/contact-us or find her on LinkedIn: linkedin.com/in/monica-ly

    EP10: AI won't break your tech. It'll break your operating model | AI ownership & operating models

  2. 21 July

    EP09: Time saved isn't the AI business case | AI use cases & business value

    Every AI pilot ends with the same slide: look how much time this will save us. Almost no one asks what happens next. Monica breaks down why "hours saved" isn't the destination of a good AI business case, it's the starting point. Using an analogy from the dashboard era (when reports got automated, nobody went home earlier — they just stopped building reports and started reading them), she unpacks why capacity created isn't the same as value delivered, and what leaders should push their teams to answer before they sign off on a use case. The test: when someone brings you a proposal built on hours saved, spend less energy checking the maths and more on the sentence that usually isn't there. What will people actually do with that time? Five minutes. One idea. Something to bring to your next leadership conversation. Have a question or topic for Monica? Visit edgered.com.au/contact-us or find her on LinkedIn: linkedin.com/in/monica-ly

    EP09: Time saved isn't the AI business case | AI use cases & business value

  3. 6 July

    EP08: Is AI actually in the job - or just in the title? | AI hiring & recruitment

    Most leaders assume an AI job title means an AI job. Often it doesn't. Michelle Rubinstein, Director at The Onset and a recruiter for over 25 years, joins Monica to unpack what's really behind the explosion in AI hiring - Head of AI roles up 28% in a year, AI Engineer roles up 37% - and the sign-off leaders keep getting wrong. Most approve headcount based entirely on their team's recommendation, without asking what business problem the hire is meant to solve. The test Michelle uses: can this person tell you how they cut costs, sped up a process, or generated revenue with AI - or do they only talk about models and prompts? If it's the latter, the salary is the cheapest part of what a bad hire costs. Five minutes. One idea. Something to bring to your next leadership conversation. Have a question or topic for Monica? Visit edgered.com.au/contact-us or find her on LinkedIn: linkedin.com/in/monica-ly

    EP08: Is AI actually in the job - or just in the title? | AI hiring & recruitment

  4. 30 June

    EP07: The Privacy Act change most businesses haven't noticed | AI Governance, Compliance, Australia

    Most conversations about the AI rules coming in December 2026 focus on whether you've got a human checking the model's work. That's the wrong question. Nhung, principal consultant at EdgeRed, joins Monica to unpack the Privacy Act's new automated decision-making obligations - and the assumption most leaders are getting wrong. If a model scores something and a human just approves whatever it recommends, that human click doesn't change the legal picture. The test that actually matters: would the outcome have changed if the model had scored it differently? If yes, the model made the decision. It also doesn't matter whether the system in question is built on AI. A computer program is a computer program under this law - your fraud-scoring spreadsheet from 2015 is just as in scope as a new AI model. Nhung walks through what's actually required, why most organisations haven't mapped where this applies to them yet, and the one question to ask your team this week before the December 2026 deadline arrives. Five minutes. One idea. Something to bring to your next leadership conversation. Have a question or topic for Monica? Visit edgered.com.au/contact-us or find her on LinkedIn: linkedin.com/in/monica-ly

    EP07: The Privacy Act change most businesses haven't noticed | AI Governance, Compliance, Australia

  5. 17 June

    EP05: You approved it - but did anyone ask a real customer? | Synthetic Research Explained

    In this episode — the first to feature a guest — Monica is joined by Oz, founder of digital transformation consultancy Bound Digital, to talk about synthetic research. That's the practice of using AI to simulate customer responses instead of running real interviews, and it's already further along than most leaders realise. Qualtrics found that 71% of market researchers expect the majority of market research to be synthetic within three years. Oz walks through where it's actually useful — early exploration, persona work, stress-testing a proposition, getting a customer journey team to a shared starting point in days rather than weeks. He also draws a clear line at the point where it stops being a shortcut and starts being a risk: when teams begin treating simulated insight as evidence for real decisions. Real people stumble, contradict themselves, change their minds, bring lived experience into a conversation. AI doesn't. Then a practical filter for working out which research questions are safe to hand to AI, and which still need a human in the room — particularly when there's real money, a real product, or a real customer outcome attached to the answer. Five minutes. One idea. Something to bring to your next meeting. Have a question or topic for Monica? Visit edgered.com.au/contact-us or find her on LinkedIn: ⁠⁠linkedin.com/in/monica-ly

    EP05: You approved it - but did anyone ask a real customer? | Synthetic Research Explained

  6. 9 June

    EP04: When to trust your gut over AI | AI Decision-Making & the Semantic Layer

    Nearly half of executives say they would override their own judgment based on AI-generated output. Most of them have no way of knowing whether that's the right call. Monica Ly unpacks why AI output looks confident whether it's right or wrong — and why that's a structural problem, not a random one. A confidently correct answer and a confidently wrong answer look identical on the slide. Unlike a human analyst who might hedge or look uncomfortable when pushed, AI won't signal doubt. That asymmetry matters when the output is informing a real decision. The structural reason most AI gets business context wrong has a name: the semantic layer. It's the difference between AI that understands what "revenue" or "active customer" actually means in your organisation, and AI that produces something that looks technically coherent but is completely wrong for your business context. Then the practical side: what to ask your team to understand whether this foundation exists, why the same scepticism that applies to numbers should apply to sources and references too, and why the executives who use AI well aren't the ones who trust it most — they're the ones who know when to push back on it. Five minutes. One idea. Something to bring to your next meeting. Have a question or topic for Monica? Visit edgered.com.au/contact-us or find her on LinkedIn: ⁠linkedin.com/in/monica-ly

    EP04: When to trust your gut over AI | AI Decision-Making & the Semantic Layer

  7. 2 June

    EP03: The pilot worked. That's not the same thing as it working | AI Pilot Scaling

    The pilot went well. The team wants to scale. The momentum is there. Most people in the room won't pause before signing off. In this episode, Monica Ly explains why a successful AI pilot is one of the more misleading signals a leader has to work with. Pilots run in carefully controlled conditions: tight scope, clean data, a small group of people watching the result closely. Production looks nothing like that. The data is messier, the attention drops off, and the edge cases someone quietly handled in the background are suddenly the system's problem. A Deloitte report this year puts a number on how few organisations successfully move AI pilots to production — and Australia is sitting further back in the pack than most leaders realise. The risk sharpens when what's being scaled isn't a dashboard but an AI agent that takes action on its own. A system that performs well under close supervision can fail in very different and more visible ways once that supervision is gone. Then the bit most boards skip: what to actually ask before approving the next round of investment, and why scaling an AI project should be treated as a different decision from approving the pilot — not just a bigger version of the same one. Five minutes. One idea. Something to bring to your next meeting. Have a question or topic for Monica? Visit edgered.com.au/contact-us or find her on LinkedIn: linkedin.com/in/monica-ly

    EP03: The pilot worked. That's not the same thing as it working | AI Pilot Scaling

  8. 21 May

    EP02: Your team is already using AI, you just can't see all of it yet. | Shadow AI & Governance

    Most leaders think they're still deciding how to roll out AI. Their teams have already started without them. Monica Ly looks at why that gap exists and what it means for the people accountable for it. Staff are using AI because it makes their job easier and faster — and a lot of them are feeding company information into tools their employer hasn't approved. Rolling out an enterprise version of ChatGPT or Claude helps, but it doesn't cover everything people reach for during the day. The result is two systems running in parallel: one you can see, one you can't. She compares it to the way critical documents still end up in personal email and download folders despite years of SharePoint policy, points to an ADAPT finding that 78% of leaders treat AI as a board-level priority while fewer than 26% have any formal governance behind it, and flags one default setting on consumer AI tools that most teams have never thought to check. Then a short, practical playbook for closing the shadow AI gap — without locking your team out of the tools that are actually making them faster. Five minutes. One idea. Something to bring to your next meeting. Have a question or topic for Monica? Visit edgered.com.au/contact-us or find her on LinkedIn: linkedin.com/in/monica-ly

    EP02: Your team is already using AI, you just can't see all of it yet. | Shadow AI & Governance

  9. 14 May

    EP01: We've seen this before — the Tableau analogy | AI Governance & Data Foundations

    People are making decisions on AI and data that they wouldn't stand behind. The pressure to move is real — but we've seen this pattern before. It just used to run slower. Monica Ly revisits the BI dashboard boom of the early 2010s, when self-service tools meant every team could publish its own version of the numbers, and the real conversation quickly stopped being about software and started being about who to trust. AI is on the same trajectory, but the barrier to entry is lower, the speed is higher, and what now sits on top of the data isn't just a chart — it's a decision, a recommendation, or an automated action. An ADAPT finding puts a number on it: Australian organisations are spending up to $28 million a year on AI projects, with 76% reporting no clear return on investment. The risk is no longer just wasted spend — in December 2026, changes to the Australian Privacy Act will require organisations to disclose when AI is being used to make automated decisions about individuals. A short playbook from the organisations that came out of the dashboard era in good shape: what they did differently, and why the same moves apply now. Five minutes. One idea. Something to bring to your next meeting. Have a question or topic for Monica? Visit edgered.com.au/contact-us or find her on LinkedIn: linkedin.com/in/monica-ly

    EP01: We've seen this before — the Tableau analogy | AI Governance & Data Foundations
  • 9 Episodes

About

A five-minute podcast for Australian business leaders navigating AI and data without a technical background. Each episode is one idea you can take into your next meeting: a question worth asking, a claim worth challenging, or a decision worth slowing down. Hosted by Monica Ly, co-founder of EdgeRed (part of The Omnia Collective) — eight years of hands on delivery experience, recorded between baby feeds and naps.

Information

  • Creator
    Monica Ly
  • Years Active
    2026
  • Episodes
    9
  • Show Website
    Nap Stack