- McKinsey executive Matthew Fitzpatrick is leaving the firm to become CEO of Invisible Technologies.
- Fitzpatrick led QuantumBlack Labs, developing AI software and tools for companies.
- Invisible Technologies, a company focused on AI training, is valued at $500M, the company said.
One of McKinsey & Company’s top executives is leaving the firm for Silicon Valley’s new promised land: the AI industry.
After 12 years at McKinsey, Matthew Fitzpatrick, senior partner and global head of QuantumBlack Labs — the software and research and development arm of the firm’s AI division QuantumBlack— is stepping down.
During his tenure, Fitzpatrick led teams that helped companies scale AI projects and oversaw the development of tools like Kedro, an open-source analytics and machine learning library that McKinsey said has been downloaded more than 17 million times since its launch in 2019.
In his next role, he will serve as the CEO of Invisible Technologies, a key company in the AI industry that has kept a relatively low profile.
Invisible Technologies specializes in data pre-training, the initial phase of training for large language model developers, and post-training, which helps refine models for companies adopting the technology. Invisible Technologies was valued at $500 million in 2024, according to a press release from the company.
Fitzpatrick said he believes Invisible can help businesses tackle one of the biggest challenges of the moment — effectively integrating AI into their operations.
“Despite the hype around AI, we’re at a point where fewer than 10% of AI models reach usage and production because enterprises don’t have the experience to evaluate, train, and operationalize them,” Fitzpatrick said in the company’s press release. “That’s where Invisible shines. I’m bullish we can help customers cross the chasm and realize the massive potential of AI.”
Fitzpatrick told BI the move came about through conversations with Invisible’s founder, Francis Pedraza, whom he met through a networking organization.
“It’s the most under-the-radar critical AI company in the US that I’ve ever seen, and it’s been involved in all of the model training for the last five years but has done very little publicity of any kind around that,” he said.
He said he sees companies having a strong demand for Invisible’s services. The challenge, however, is “not growing too fast that we in any way sacrifice quality,” he said.
Somesh Khanna, a former colleague of Fitzpatrick’s, told BI Fitzpatrick built a reputation for changing McKinsey’s talent pool, bringing more engineers and employees with quantitative skills into the fold. McKinsey told BI that it now employs 7,000 people as technologists, designers, and product managers. Fitzpatrick was responsible for overseeing 1,000 of them, a rep for Fitzpatrick told BI.
“The biggest thing is that in the McKinsey model of the past it was very hard for atypical profiles such as PhDs or data engineers or scientists to integrate into the culture of the firm. McKinsey’s philosophy centered around acquiring amazing talent and teaching these individuals the McKinsey approach to problem-solving and client service,” Khanna, a senior partner who retired from the firm in May after almost three decades, told BI.
“These guys, the new team that Matt was hiring and developing, were even more different — people who basically were hands-on keyboards guys, not power pointers.”
Khanna said Fitzpatrick was critical in integrating this new breed of talent with deep technical skills into the firm.
Read the full article here
