Host Paul Bibby introduces the Orchestrating AI podcast for senior leaders and is joined by Sydney-based AI executive and strategist Nick Munro, who explains that AI failures are typically organisational - skipped fundamentals, unclear problem definitions, weak data readiness, poor change management, and inadequate governance - rather than technical. They discuss a five-layer capability approach spanning strategy, infrastructure, governance, people/adoption, and measurement/telemetry, emphasising taking a thin slice across all layers to move from experiments to proofs of concept, pilots, and scalable “product-like” AI services with reusable patterns. Munro highlights risks such as uncontrolled generative outputs, shadow IT, and public-model data leakage, and argues governance is a board-level responsibility. Practical ROI examples include high-volume operational document processing, cost-optimised inference, behavioural collections uplift, and near-real-time contact-centre coaching. Munro’s advice: start with business problems, invest in change management, build cross-functional governance early, and prioritise clean data and measurable outcomes.

Episode Downloads

The Five Layer Model - https://forms.office.com/e/XK8qsFBVQD

FREE AI Readiness Assessment - https://forms.office.com/e/3dwbqqdJFj

Chapters

00:00 Welcome and Mission

02:29 Why AI Falls Short

04:17 Framework For Scaling AI

04:32 Framework In Practice

06:46 Chatbot Risk Example

09:06 From Pilots To Scale

14:50 Owning The AI Agenda

17:35 Skills Culture Shift

21:07 Risks Noise And Shadow IT

24:30 Public Vs Private Data Risk

25:30 Data Residency Risks

26:02 Preventing Data Leaks

26:13 Safe Internal AI Tools

26:59 High ROI Use Cases

27:33 Scaling Inference Costs

28:57 Scale Over Shiny

30:07 Agentic AI Explained

32:29 Grocery Agent Reality Check

34:50 Why Agents Cost More

39:05 Real Time Call Coaching

42:50 Board Level Governance

47:45 Guest Signature Question - GSQ

50:54 Host Takeaways Wrap

Disclaimer: The opinions and views shared in the Orchestrating AI Podcast are those of the individual speakers. They do not represent the views of their employers, organisations, or any affiliated bodies. The discussions in this podcast are for general information only and are not intended as professional, technical, financial, or legal advice. Listeners should always seek appropriate advice before making decisions based on the topics discussed.

Host Paul Bibby discusses AI governance and regulation with Siyanbola Feyisetan (Nayis), an applied AI and digital transformation leader focused on building explainable, auditable, dependable systems. They argue governance is a leadership responsibility because accountability for AI decisions cannot be outsourced, though IT, legal, risk, and compliance implement and monitor controls. Feyisetan outlines global regulation (including the EU AI Act, sovereign AI approaches, and fragmented US state rules) and highlights the EU AI Act’s risk-based categories. For organisations operating across jurisdictions, she stresses complying with each region’s laws. They describe practical governance elements: integrated policies on data, model use, risk classification, testing/monitoring, human oversight for high-impact decisions (hiring, credit, healthcare), documentation, approvals, audit trails, and minimum viable governance via clear usage rules, lightweight risk checklists, approval paths, staff training, and cross-functional reviews.

Downloads

Minimum Viable AI Governance Checklist - https://forms.office.com/e/4cVw3LvQB8 FREE AI Readiness Assessment - https://forms.office.com/e/3dwbqqdJFj

Chapters

00:00 Accountability Starts Here

01:24 Meet Nayis Feyisetan

02:30 Leadership Owns Governance

04:54 Global AI Regulation Tour

08:51 Building Governance Structure

10:45 Policies Guardrails Oversight

13:13 Documentation Audit Trails

14:53 Real World Risk Scenarios

16:58 Bias Data And Feedback

20:23 Minimum Viable Governance

25:32 Signature Question Lessons

28:32 Key Takeaways And Wrap

Disclaimer:

The opinions and views shared in the Orchestrating AI Podcast are those of the individual speakers. They do not represent the views of their employers, organisations, or any affiliated bodies.

The discussions in this podcast are for general information only and are not intended as professional, technical, financial, or legal advice. Listeners should always seek appropriate advice before making decisions based on the topics discussed.

Paul Bibby from Orchestrato introduces Orchestrating AI, a thought-leadership podcast focused on bringing clarity, credibility, and real-world insights into Artificial Intelligence in business. The show features open, practical conversations with guests spanning AI strategy, data, engineering, and automation, sharing lessons on building, deploying, and scaling AI across organisations in the UK, USA, and beyond. Episodes explore the full lifecycle of AI adoption with the aim of making AI understandable, actionable, and achievable for business leaders, and each ends with the signature question about what guests would do differently if starting their AI journey today. The script also includes a disclaimer that views are individual and not professional or legal advice.

00:00 Show Intro

00:21 What To Expect

00:45 Mission And Signature Question

01:01 Like And Subscribe

01:08 Disclaimer

Disclaimer:

The opinions and views shared in the Orchestrating AI Podcast are those of the individual speakers. They do not represent the views of their employers, organisations, or any affiliated bodies.

The discussions in this podcast are for general information only and are not intended as professional, technical, financial, or legal advice. Listeners should always seek appropriate advice before making decisions based on the topics discussed.