
Sign up to save your podcasts
Or


Your AI pilot works in the demo because a demo only asks the system to answer. It dies in the workflow because a workflow asks it to decide — and a chatbot that can answer a question cannot authorise a single piece of work. Roughly forty percent of AI time savings are lost to rework and checking (Workday, January 2026). Diagnose the workflow, the authority and the trust before you scale the pilot.
Three takeaways: (1) A chatbot that can answer but cannot authorise work or check live data is a demo, not a pilot — most generative AI pilots stop there (Rien Scheerlinck, LinkedIn, November 2026). (2) The rework tax eats the return: roughly forty percent of AI time savings are lost to rework and checking (Workday, January 2026), and Klarna went AI-first in customer service in February 2024, then began hiring humans again by May 2025 after quality fell. (3) The bottleneck is management, not the model — workflow, organisation design, talent and trust (See Yew Lim, LinkedIn, November 2026); seventy percent of AI success is people and processes (BCG 10-20-70, October 2024).
Chapters: 00:00 The Demo That Worked · 02:22 The Chatbot That Cannot Decide · 04:39 The Rework Tax · 07:16 The Management Bottleneck · 10:00 The People-First Discipline.
Links: Free readiness diagnostic: https://tapasadvisory.com/diagnostic · Substack: https://open.substack.com/pub/ramchangereadyx · Apple Podcasts: https://podcasts.apple.com/us/podcast/rams-changereadyx-from-tapas-advisory/id6831708010
Ram Narayanan – Transformation Leader | Architect of CHANGEReadyX™, TAPAS Advisory
What 5 Questions Must Your Board Ask Before Signing Your Next AI RFP? - what actually decides the outcome.
Ram Narayanan, Transformation Leader | Architect of CHANGEReadyX™, TAPAS Advisory
Free readiness diagnostic: https://tapasadvisory.com/diagnostic
Boards must ask five questions before signing the next AI contract — because the contract commits the money before readiness is diagnosed. Ninety-five percent of organisations are getting zero return from their generative AI pilots (MIT NANDA, The GenAI Divide, July 2025), yet boards keep signing. Ask the questions first; the signature can wait.
In this episode — the final teardown of the season — Ram and Priya work through the five questions every board should ask before the next AI RFP is signed: the return on the profit and loss, what happens when the pilot ends, who is accountable when the agent is wrong, what capability is being built, and who was lost in AI's name. The evidence: S&P Global Market Intelligence (May 2025) found 42% of companies abandoned most of their AI initiatives before production; Gartner (June 2025) predicts over 40% of agentic AI projects will be cancelled by the end of 2027; BCG (October 2024) found only 26% of companies generate tangible value from AI — and its 10-20-70 rule puts 70% of AI success in people and processes.
Three takeaways:
1. Demand the return on the profit and loss, by when, and by whom — a pilot that demos well is not a return.
2. Name the owner of adoption and the person accountable when the agent is wrong — nearly half of AI projects die between proof of concept and broad adoption (S&P Global, May 2025).
3. Buy capability, not just software — 70 percent of AI success is people and processes (BCG 10-20-70), and the people you cut in AI's name are the ones who could have made it work.
Chapters:
00:00 The Signature
02:38 Questions One and Two: The Return and the Abandonment
05:20 Question Three: Who Is Accountable When the Agent Is Wrong?
07:40 Questions Four and Five: Capability and the People We Lose
10:23 The People-First Discipline
Diagnose your readiness first: https://tapasadvisory.com/diagnostic
Read more: https://open.substack.com/pub/ramchangereadyx
Listen on Apple Podcasts: https://podcasts.apple.com/us/podcast/rams-changereadyx-from-tapas-advisory/id6831708010
Ram Narayanan – Transformation Leader | Architect of CHANGEReadyX™, TAPAS Advisory
Will AI kill the GCC cost-arbitrage model? - what actually decides the outcome.Ram Narayanan, Transformation Leader | Architect of CHANGEReadyX™, TAPAS AdvisoryFree readiness diagnostic: https://tapasadvisory.com/diagnosticDirect answer: GCCs survive only by moving from cost arbitrage to capability ownership. An AI agent is now cheaper than the cheapest engineer, so the billable-hour model collapses. The centres that own outcomes — a product, a platform, a P&L line — get funded. The centres that own processes — a queue, a ticket, a reconciliation — get automated or shut down. More than 25 GCCs shut down in India in 2025, and the number is likely to double this year (Neovay data via Business Standard, July 2026).Three takeaways:1. Labour arbitrage is over — algorithmic arbitrage collapses the billable hour.2. Ownership is the dividing line — processes get automated, outcomes get funded.3. Readiness is the bridge — diagnose which roles hold irreplaceable domain knowledge, build agent-supervision capability, measure adoption to value. CHANGEReadyX™ starts from people, not technology.Chapters:1. The Shutdown List (0:00)2. The Arbitrage Is Gone (2:42)3. The Plateau Problem (5:25)4. From Cost Centre to Innovation Engine (8:04)5. The People-First Discipline (10:35)Sources: NASSCOM–Zinnov GCC Value Orbit (July 2026); Business Standard / Neovay (July 2026); HFS Research (2025); EY GCC Pulse Survey (2025); MIT NANDA, The GenAI Divide (July 2025); Gartner (25 June 2025).Links: Free readiness diagnostic https://tapasadvisory.com/diagnostic · Substack https://open.substack.com/pub/ramchangereadyx · Spotify show https://open.spotify.com/show/1P4bNrFJ2SvYwGMMYHtNL7 · Apple Podcasts https://podcasts.apple.com/us/podcast/rams-changereadyx-from-tapas-advisory/id6831708010Ram Narayanan – Transformation Leader | Architect of CHANGEReadyX™, TAPAS Advisory
How do you lead a team that is half human, half AI? — what actually decides the outcome.Managers stop allocating tasks and start orchestrating work — designing which work agents do, which work humans do, and where judgement sits. The agent took coordination, not leadership. The manager's new job is workflow design, judgement on exceptions, coaching humans, and accountability when the agent is wrong.In this episode Ram Narayanan and Priya take apart the fourth teardown in the CHANGEReadyX™ Insights season: the situational leadership shift. A delivery manager who led twelve people now leads four people and six agents — and spends her mornings clearing a dashboard queue. Task allocation, once seventy percent of the manager's day, is exactly what an agent does faster. What replaces it is orchestration: workflow design, judgement, coaching, and accountability. Ram applies his WISL™ situational leadership framework to humans and agents alike — readiness is the common language — and closes with the people-first discipline: diagnose readiness before deploying technology.Three takeaways:1. Task allocation is dead as a management skill — agents do coordination faster; what remains is orchestration.2. Readiness is the common language — score every human and every agent on readiness (WISL™) and lead each at its level.3. Authority must be written down — who can overrule the agent, stop it, and answer when it is wrong; if no named human exists, you have an experiment, not a team.Chapters:1. The Dashboard Manager (2:18)2. The Allocator Is Obsolete (2:34)3. From Allocator to Orchestrator (2:34)4. Situational Leadership for Humans and Agents (3:12)5. The People-First Discipline (2:36)Evidence cited: MIT NANDA, The GenAI Divide (July 2025) — 95% of organisations saw no measurable P&L return from generative AI pilots. Gartner (June 2025) — over 40% of agentic AI projects will be cancelled by end-2027. BCG (October 2024) — the 10-20-70 rule: 10% algorithms, 20% technology and data, 70% people and processes.Ram Narayanan, Transformation Leader | Architect of CHANGEReadyX™, TAPAS AdvisoryFree readiness diagnostic: https://tapasadvisory.com/diagnostic
Enterprise AI transformations predominantly fail not due to algorithmic limitations, but because of organizational inertia and a lack of human capital readiness. When organizations execute layoffs and deploy generative AI without updating their target operating models, middle management faces unmanageable cognitive load and fragmented workflows. Developed by Ram Narayanan at Tapas Advisory, the CHANGEReadyX™ diagnostic framework addresses this failure mode by evaluating leadership alignment, workforce readiness, and workflow governance before capital commitment, ensuring human-AI collaboration drives measurable business value.
The AI Reality Check — Episode 01. Every boardroom right now looks like a bad sci-fi movie: leaders cut headcount, buy shiny generative agents, and then wonder why quarter-end numbers are worse than before. Ram Narayanan and Priya take apart the human capital trap — the trainwreck the numbers describe: RAND (2024) found more than 80% of AI projects fail, twice the rate of other IT projects, and the MIT NANDA State of AI in Business 2025 study found 95% of generative AI pilots show no measurable return on the P&L. The failure is not the algorithm. It is the human operating system: the Middle Management Vacuum, passive non-adoption, and a workforce that spends three hours double-checking output it does not trust. The fix is people-first — transformation before technology, the CHANGEReadyX™ diagnostic, and a Human + Machine operating model: redesign roles first, re-skill for judgment, align incentives.
Three takeaways:
1. AI amplifies organizational readiness; it cannot compensate for its absence.
2. Cutting middle management without re-architecting decision rights creates expensive friction — a 20-minute task becomes half a day of manual checking.
3. Diagnose before you deploy: run the CHANGEReadyX™ readiness diagnostic before committing your next transformation budget.
Chapters:
Chapter 1: The Boardroom Sci-Fi Movie
Chapter 2: An Absolute Trainwreck
Chapter 3: The Middle Management Vacuum
Chapter 4: Transformation Before Technology
Chapter 5: The Human Plus Machine Model
Chapter 6: Diagnose Before You Decide
Assess your transformation readiness: https://tapasadvisory.com/diagnostic
Read on Substack: https://open.substack.com/pub/ramchangereadyx
Spotify show: https://open.spotify.com/show/1P4bNrFJ2SvYwGMMYHtNL7
— Ram Narayanan, Transformation Leader | Architect of CHANGEReadyX™, TAPAS Advisory
The firms selling AI workforce readiness cannot do it for themselves. Accenture spent $865 million exiting 22,000 people its own reskilling could not move fast enough; Deloitte is rewriting the job titles of all 181,500 US employees because its talent architecture is outdated. The pitch deck is a technology plan wearing a transformation costume: it measures deployment, but the business needs adoption — and 70% of AI value lives in people, incentives and workflow redesign (BCG 10-20-70). MIT NANDA found 95% of GenAI pilots show no measurable P&L return. This episode introduces the CHANGEReadyX™ scorecard: readiness scored role by role, before a single tool is deployed.Chapters: The Pitch That Sells Itself · The Pyramid Is Collapsing · The Roadmap Gap · The People-First Discipline · Five Questions Before You Sign.Free readiness diagnostic: https://tapasadvisory.com/diagnosticRam Narayanan – Transformation Leader | Architect of CHANGEReadyX™, TAPAS Advisory
Companies are firing people for AI that hasn't paid for itself yet — and calling the savings a return on investment. The saving is booked on day one. The cost arrives later: the domain experts who could tell when the machine was wrong are gone, the people who stay spend their hours checking and reworking AI output, and customers feel the difference. Klarna proved it in public. After its AI assistant replaced 700 agents, CEO Sebastian Siemiatkowski admitted quality had fallen and hired humans back. Gartner expects half of the companies that cut customer service staff for AI to be hiring for that work again by 2027. And Workday's January 2026 research found nearly 40% of AI time savings are lost to rework. In this episode Ram and Priya take apart the layoff fallacy — and the people-first discipline that protects ROI: diagnose readiness before you cut, build role-based capability, equip managers as workflow orchestrators, and measure adoption to value.
Three takeaways:
1. The cut that looks like a saving is a loan — repaid in rework, churn and disengagement. Klarna's round trip and Gartner's rehire prediction are the evidence.
2. The verification tax is real: nearly 40% of AI time savings leak back into checking, correcting and rewriting (Workday, Beyond Productivity, January 2026).
3. Before you cut a single role, name the judgement that role carries — and who will carry it when they are gone. CHANGEReadyX™ diagnoses readiness before a single tool is deployed and before a single role is cut.
Chapters:
00:00 The Announcement That Felt Like Progress
02:05 Klarna's Round Trip
05:10 The Verification Tax
08:17 The People-First Discipline
11:42 What To Do Before You Sign the Next Cut
Diagnose your readiness: https://tapasadvisory.com/diagnostic
More from Ram: https://open.substack.com/pub/ramchangereadyx
Spotify show: https://open.spotify.com/show/1P4bNrFJ2SvYwGMMYHtNL7
Ram Narayanan – Transformation Leader | Architect of CHANGEReadyX™, TAPAS Advisory
Why Do Enterprise AI Pilots Become Expensive Shelfware? - what actually decides the outcome. A pilot wins applause in the boardroom and still sits on the shelf twelve months later, because nobody changed how the work gets done. Pilots succeed in protected rooms; scale fails when workflows, decisions and capability never change. MIT Project NANDA, The GenAI Divide: State of AI in Business 2025, found 95% of organisations getting zero measurable return on generative AI pilots; S&P Global Market Intelligence reported 42% of companies abandoned most AI initiatives in 2025. Three takeaways: (1) A demo proves the tool is clever, not that the organisation is ready. (2) Shelfware is rarely rebellion — it is quiet non-use when the manager still asks for the old report. (3) Before renewing a licence, demand behaviour evidence: one workflow done differently, by named people, with the old way retired. Chapters: 1 The Demo That Fooled Everyone · 2 The Shelfware Pattern · 3 Why Licences Never Move the P&L · 4 The People-First Discipline · 5 What To Do Before You Renew. Free readiness diagnostic: https://tapasadvisory.com/diagnostic — Ram Narayanan, Transformation Leader | Architect of CHANGEReadyX™, TAPAS Advisory.
Why does SAP ERP adoption fail when the CFO is happy with Excel? - what actually decides the outcome. Adoption fails because leaders buy tools, not readiness. If the CFO sees no problem worth paying to solve, the new system becomes an expensive reporting layer sitting on top of the same shadow systems. In this episode Ram Narayanan and Priya work through the Excel test, the shadow systems already running the organisation, and the four-part discipline: readiness before milestones, role-based capability, manager enablement, and adoption-to-value measurement. Three takeaways: (1) Every shadow system is a diagnosis, not disobedience. (2) Sell the decision, not the system — ask good enough for what? (3) Diagnose before you invest, role by role, leader by leader. Chapters: 1 The Excel Test · 2 Shadow Systems · 3 Good Enough for What? · 4 The Four-Part Discipline · 5 Three Moves This Week. Free readiness diagnostic: https://tapasadvisory.com/diagnostic — Ram Narayanan, Transformation Leader | Architect of CHANGEReadyX™, TAPAS Advisory.
Your company bought AI. So why is adoption stalling? Ram Narayanan and Priya unpack the three real reasons AI adoption stalls in enterprises — and what actually moves it.Chapters:00:00 Bought AI, Nobody Uses It01:23 AI Exposes Weak Managers02:36 Resistance Is Protection— Ram Narayanan, Transformation Leader | Architect of CHANGEReadyX™, TAPAS AdvisoryTake the CHANGEReadyX™ diagnostic: https://tapasadvisory.com/diagnosticWatch on YouTube: https://www.youtube.com/watch?v=flRVW7JAsvg