Matt Fitzpatrick - CEO, Invisible. Tracked across 23 mentions in podcasts and expert conversations analyzed by Sonic.

Matt Fitzpatrick

CEO of Invisible, advocating for human-in-the-loop AI integration services.

Claims

As speaker

Mentioned in

Key positions and views

Out-of-the-box enterprise AI agents are largely ineffective, with accuracy rates as low as 33% on complex tasks, necessitating deep, external integration.

Externally-driven AI development builds are twice as effective as those managed by internal enterprise teams, advocating for an outsourced, expert-led model.

The belief that synthetic data will replace the need for human feedback in AI within the next 2-3 years is 'fundamentally flawed.'

A shift from remote to in-person work can yield an 'exponential increase in productivity,' justifying the establishment of multiple physical offices globally.

The high failure rate of enterprise AI projects, combined with high consumer adoption of GenAI, represents a massive market opportunity for companies that can bridge the gap.

Timeline

First 9 years

Fitzpatrick's company, Invisible, operated with minimal funding, having raised only $7 million in primary capital.

Recent Past

Invisible shifted its operational model from fully remote to primarily in-person, a move Fitzpatrick claims led to an exponential increase in productivity.

Recent Funding Round

The company secured a significant new funding round, raising $130 million, marking a dramatic change in its financial trajectory.

Last 45 Days

Fitzpatrick reports a surge in business development, with Invisible securing 12 new enterprise deals in this short period.

Current

The company employs 450 people and is investing heavily in technology, leading to a decision to forgo profitability for the current year in favor of growth.

Podcast consensus on Fitzpatrick

Points of consensus

▶Fitzpatrick consistently argues that enterprise AI projects are failing at a high rate, citing reports from MIT and Gartner to support the view that out-of-the-box solutions are ineffective without significant customization.Mar 2026

▶He emphasizes the critical role of human-in-the-loop processes and specialized expertise for successful AI deployment, as evidenced by his company's 1.3 million expert marketplace and collaborations with entities like the U.S. Navy.Mar 2026

▶A core belief is that the market for AI integration and services is massive and growing, drawing parallels to the rise of IT services firms like Accenture that were built to customize enterprise software.Mar 2026

▶He presents his company, Invisible, as experiencing rapid, recent growth, highlighting a new $130 million funding round after years of minimal investment, a surge in enterprise deals, and a significant increase in productivity.Mar 2026

Points of debate

▶Fitzpatrick's narrative presents a sharp contrast between Invisible's first nine years, where it raised only $7 million, and its recent success in raising $130 million and securing numerous enterprise deals, suggesting a significant strategic pivot or market shift.Mar 2026

▶He champions a transition from a fully remote to an in-person work model, claiming it led to an 'exponential increase in productivity,' a stance that contrasts with the prevailing remote-first trend in the tech industry.Mar 2026

▶He presents a conflicting view of the AI market, simultaneously highlighting high consumer adoption (60% weekly use per KPMG) and extremely low enterprise success (5% operational deployments per MIT), framing the gap as his company's primary opportunity.

▶Fitzpatrick argues against the idea that synthetic data will replace human feedback within 2-3 years, positioning his human-centric data labeling model as a durable necessity against a prevailing tech narrative.Mar 2026

Key themes

▶The Failure of 'Out-of-the-Box' AI

Fitzpatrick consistently argues that generic, off-the-shelf AI solutions are failing enterprises. He cites reports from Salesforce AI Research, MIT, and Gartner to demonstrate low accuracy rates, high project cancellation predictions, and a mere 5% operational deployment rate for generative AI projects.

This theme suggests a significant market opportunity for service-based companies that can provide the complex integration, customization, and 'forward-deployed' engineering required to make enterprise AI functional and effective.

▶Invisible's Strategic Pivot and GrowthMar 2026

The narrative details a dramatic shift for the company Invisible, moving from a slow-growth, minimally funded entity to a rapidly expanding firm with $130 million in new funding. This growth is attributed to a new go-to-market strategy, a shift to in-person work, and securing major enterprise and government contracts.

Investors should note the company's high-burn growth strategy, as Fitzpatrick states Invisible will not be profitable this year due to heavy technology investments, indicating a focus on capturing market share over short-term financial performance.

▶Human-in-the-Loop as a Core DifferentiatorMar 2026

Fitzpatrick positions human expertise as essential for successful AI, directly challenging the notion that synthetic data will suffice. He highlights Invisible's marketplace of 1.3 million experts, its ability to deploy 26,000 experts weekly, and its work fine-tuning models for complex tasks like controlling drone swarms for the U.S. Navy.

This focus on a massive, flexible human workforce as a complement to AI technology is a key differentiator that could provide a defensible moat against purely software-based competitors.

▶Citing External Data to Build Authority

Throughout his discussion, Fitzpatrick frequently references reports and statistics from authoritative third-party sources like MIT, Gartner, KPMG, Salesforce, and the World Economic Forum. This technique is used to validate his market analysis, underscore the problems his company solves, and lend credibility to his strategic vision.

Analysts should verify these cited reports to understand the full context, as the specific framing of these statistics is crucial to Fitzpatrick's narrative about the state of the AI industry and his company's role within it.

Source episodes

Suggested prompts

How does Fitzpatrick's emphasis on 'forward-deployed engineers' compare to the business models of traditional IT services firms like Accenture and Cognizant? &nearr;What specific changes in Invisible's business model or the market enabled its recent $130 million funding round after nine years of minimal investment? &nearr;How sustainable is a business model that relies on a marketplace of 1.3 million human experts in an era of increasing automation and synthetic data generation? &nearr;To what extent does the claimed 'exponential increase in productivity' from switching to in-person work reflect company-specific factors versus a universally applicable principle for tech firms? &nearr;

Key concepts

Invisible (company) 11 ep Enterprise AI Deployment 6 ep Market Research & Reports 9 ep Business Strategy 4 ep AI Model Accuracy 3 ep Data Labeling & Human Feedback 3 ep Venture Capital & Funding 2 ep Remote vs. In-Person Work 2 ep

Notable quotes

“Salesforce AI Research released this report that if you take, if you test a lot of the out-of-the-box agents on single-term and multi-term workflows, they're about 58% accurate on single-term and 33% accurate on multi-term workflows, which means they don't really work.”

Matt · Matt Fitzpatrick: Who Wins the Data Labelling Race & Why Al Needs Forward-Deployed Engineers

“historically, Invisible had only raised $7 million of primary capital in its entire nine-year journey. We've now, we initially announced $100 million. Actually, right now we're raised $130 million.”

Matt · Matt Fitzpatrick: Who Wins the Data Labelling Race & Why Al Needs Forward-Deployed Engineers

“MIT just released this report that 5% of Gen AI deployments are working in any form.”

Matt · Matt Fitzpatrick: Who Wins the Data Labelling Race & Why Al Needs Forward-Deployed Engineers

“You've seen Gartner saying 40% of enterprise projects will likely be canceled by 2027.”

Matt · Matt Fitzpatrick: Who Wins the Data Labelling Race & Why Al Needs Forward-Deployed Engineers

Report last updated: Jul 28, 2026

Create a free account to see Matt Fitzpatrick's full intelligence report - every claim, the relationship network, and AI Q&A across all sources. No card needed.

Get started free

Back to Entities Intelligence Report