Artificial intelligence Robotics Startups Venture

Insight Partners Has Invested More Than $4B Into AI. What’s Next?

Illustration of paper money with Insight Partners overlaid on it.

Editor’s note: This is part of a series in which we interview active investors in artificial intelligence. Previous interviews were with investors at General Catalyst, Bessemer Venture Partners and Accel.  

New York-based investment firm Insight Partners has invested around $4 billion in artificial intelligence companies over the past five years.

That’s according to George Mathew, a partner at the firm who has led some of its key investments in the AI sector.

George Mathew, partner at Insight Partners

“Every business is having a transformative moment,” Mathew said in a conversation with Crunchbase News, speaking to the new wave of innovation unleashed by the launch of ChatGPT late last year. Such generative AI models have raised the bar that will “reformat all of software,” he said.

Mathew joined Insight in February 2021 and made his very first investment at the firm leading the Series B in Weights & Biases, a machine-learning operations company used by OpenAI for tracking its transformer models. Seven months later, Weights & Biases went on to raise a Series C that valued it as a new unicorn.

Insight was already one of the most active startup investors in 2021 and into the first half of 2022, based on an analysis of Crunchbase data. The firm added 146 portfolio companies in 2022 alone, for a total of 750 primary investments with $50 billion in capital commitments, according to the firm.

Search less. Close more.

Grow your revenue with all-in-one prospecting solutions powered by the leader in private-company data.

The firm reportedly has $10 billion in dry powder, but is said to be scaling back on the $20 billion target size of fund 13 and deploying funds over a longer time horizon. Its funding pace in 2023 has slowed significantly from its peak in Q1 2022, per Crunchbase.

Most of Insight’s funding to its AI portfolio companies has come from its recent funds. Those are fund 11, a $9.5 billion fund announced in 2019, and fund 12, announced in 2022 and the firm’s largest at $20 billion (at a time when private company financings generally slowed).

Beyond the data stack

Insight is tackling AI investments at three distinct layers. They include the modern data stack, MLOps (machine-learning operations) and generative AI applications.

The firm led the Series C in data orchestration platform Astronomer in 2022.  However, the data stack is pretty well invested at this time, Mathew said. “A lot of our focus definitely shifted toward this next generation of machine-learning operations and specifically how LLMs are being built,” he said.

The acquisition of MosaicML by Databricks is indicative of a data stack company becoming a key player in MLOps, he said.

Mathew’s other investments in AI include data observability platform Acceldata, AI model performance management company Fiddler AI, and synthetic data creation company

Other investments in MLOps include data collaboration company Atlan, customer data platform RudderStack, and analytics company Kubit, which is built on top of Snowflake and Databricks.

Generative AI apps

As for what he’s looking for now as he seeks out generative AI companies to invest in, Mathew said it’s “companies that are building domain-specific models on top of private data sets with great user experiences and workflows.”

On the applications side, the firm is interested in vertical opportunities in financial services and health care as well as software — namely robotic process automation — that will be reimagined by this technology. Two portfolio companies reinventing robotic process automation with LLMs and transformers are Bardeen and Workato.

Insight is also interested in horizontal opportunities like Jasper, for which it led the $125 million Series A — a month before the launch of ChatGPT. The firm also invested in entertainment dubbing service deepdub, which translates not only voice but tonality into 40 different languages. And it led the Series A for Hour One, which provides an avatar built by a generative model for video communications.

Related Crunchbase Pro query

Illustration: Dom Guzman

Stay up to date with recent funding rounds, acquisitions, and more with the Crunchbase Daily.

Copy link