AI and the Future of Work: Artificial Intelligence in the Workplace, Business, Ethics, HR, and IT for AI Enthusiasts, Leaders and Academics

388: From AI Hype to Real Deployment: What Enterprise Leaders Keep Getting Wrong, with Matt Fitzpatrick, CEO of Invisible Technologies

10 snips

May 11, 2026

Guest

Matt Fitzpatrick

Matt Fitzpatrick, CEO of Invisible Technologies and former Global Head of QuantumBlack Labs at McKinsey, explains why enterprise AI lags despite model progress. He discusses messy legacy data, the importance of human feedback, redesigning workflows rather than layering AI, and why business ownership and clear output standards are crucial for real deployments.

Ask episode

AI Snips

Chapters

Transcript

Episode notes

INSIGHT

Model Progress Outpaces Enterprise Adoption

Model performance has improved dramatically but enterprise adoption lags because deployment requires more than algorithms.
Matt Fitzpatrick cites a decade of experience: models up 60–70% yet organizations struggle with adoption due to change management and incentives.

ADVICE

Cut Meetings To Boost Learning Velocity

Reduce decision velocity and meeting inertia to increase engineering throughput and learning velocity.
Matt built almost no standing meetings and pushes teams to spend time with customers and building products to accelerate feedback loops.

ANECDOTE

How Invisible Rebuilt Inventory Forecasting

Invisible runs four modular platforms: Neuron for data, Atomic for processes, Axon for agents, and Synapse for evaluation, plus a 1.1M expert marketplace.
Example: they rebuilt Swiss Army luggage forecasting by generating ~700 tables and doubling SKUs accurately forecasted.

Get the Snipd Podcast app to discover more snips from this episode

Get the app

Send us Fan Mail

Matt Fitzpatrick is the CEO of Invisible Technologies, an AI platform used to improve models for more than 80% of the world’s leading AI companies, including Microsoft, AWS, and Cohere. The company has raised $100 million and scaled to $134 million in revenue, making it one of the fastest-growing AI companies globally.

Before joining Invisible, Matt was the Global Head of QuantumBlack Labs at McKinsey, where he led large-scale AI and data engineering efforts and helped enterprises move from experimentation to production.

In this episode, Matt draws on years spent inside enterprise AI deployments to challenge the gap between model progress and real-world adoption, and to explain why most organizations still struggle to turn AI into measurable business outcomes.

In this conversation, we discuss:

Why enterprise AI adoption lags far behind model performance improvements, and why most organizations still struggle to turn technical progress into real business impact
The hidden role of messy, fragmented legacy data, and why decades of accumulated systems make it nearly impossible to deploy reliable AI at scale
Why defining “good” output in generative AI is far harder than expected, and how unclear standards stall deployment across high-stakes enterprise workflows
The case for redesigning workflows from scratch, and why layering AI on top of existing processes fails to create meaningful efficiency gains
Why most AI initiatives fail due to lack of business ownership, and how separating technology teams from operators prevents projects from reaching production
How fear-driven narratives about job loss are slowing adoption, and why AI is more likely to shift work toward higher-value tasks than eliminate roles entirely

Explore this conversation:

00:00 Intro and Fun Fact

03:57 Matt Fitzpatrick's Path From McKinsey to Invisible Technologies

09:56 Scaling Enterprise AI with Modular Platforms and Clean Data

12:44 The Crucial Role of Expert Human Feedback in Model Training

17:56 Why 95% of Enterprise AI Projects Never Reach Production

21:38 The Missing Link: Why True AI Transformation Requires Business Ownership

26:54 Overcoming AI Fear and the Reality of Jevons Paradox

32:24 Responsible AI: Governing Outcomes Over Technology

39:05 The Future of Work: Moving From Administration to Innovation

44:12 Where to Connect with Matt Fitzpatrick and Invisible Technologies

Resources: