Intrepid Growth Partners Closes $525M AI Growth Fund

Intrepid Growth Partners has formally launched with the final close of its inaugural $525M Fund I. The Toronto- and London-based firm is directing the vehicle toward growth-stage companies using machine intelligence to redesign industries, with a particular focus on the Canada-UK corridor as well as the United States and Europe.

The fund is backed by more than 80 global limited partners, including Temasek, Abu Dhabi Investment Council, British Business Bank, Export Development Canada, and Business Development Bank of Canada. The institutional roster matters because Intrepid is selling a specific investment judgment: as AI makes prediction cheaper, value should concentrate around judgment, proprietary data, distribution, workflow redesign, and the companies able to reorganize a system around those assets.

Fund I had already invested in 9 companies before the formal launch. Intrepid's portfolio and operating footprint put the $525M close behind a strategy already in motion rather than a fund waiting for its first proof of activity.

What Intrepid Growth Partners Closed

Intrepid Growth Partners announced the final close of Fund I on September 14, 2026, alongside the firm's formal launch. The vehicle closed with $525M in committed capital. That wording matters: this is capital committed to an investment fund, not company revenue, a valuation, or an amount already deployed.

The final close completes a fundraising arc that was publicly described in January 2025 as a planned $500M growth fund. The official launch release now identifies the finished vehicle, its close state, and its institutional LP base. It does not disclose individual LP commitment sizes, fund economics, total capital deployed, or performance.

The named LPs span sovereign, government-backed, and institutional capital. Alongside Temasek, the release identifies Abu Dhabi Investment Council, British Business Bank, Export Development Canada, and Business Development Bank of Canada. More than 80 LPs participated in total, alongside other US and global institutions and family offices.

The Team Behind Fund I

Intrepid was founded in 2023 by Mark Machin, Mark Shulgan, and Ajay Agrawal. Their backgrounds explain why the firm talks about machine intelligence through both institutional capital and operating economics.

Mark Machin, Co-founder and Managing Partner, previously served as President and CEO of CPPIB after more than 2 decades at Goldman Sachs. Mark Shulgan, Co-founder and Partner, built and led growth-equity platforms at CPPIB and OMERS. Ajay Agrawal, Co-founder and Partner, is a University of Toronto professor, founder of Creative Destruction Lab, and co-author of Prediction Machines and Power and Prediction. The current roles and biographies are detailed on Intrepid's team page.

That combination gives the new manager 3 different but connected views of the same problem: how institutions allocate capital at scale, how growth investors underwrite companies moving into larger markets, and how cheaper machine prediction changes the economics around decision-making.

Why Intrepid Is Focused on System Redesign

Intrepid's thesis is narrower than a general bet on AI adoption. The firm says it is looking for companies with demonstrated commercial traction that can rebuild how an industry works, rather than add a single AI feature to an existing task.

Agrawal's economic frame helps make the distinction concrete. When prediction becomes cheaper, the scarce complements become more valuable: judgment, proprietary data, distribution, workflow redesign, and the ability to reorganize a system. Fund I is designed to back businesses that can own more of those complements as AI moves from demonstration into daily operations.

That is also why the firm's creation statement emphasizes founders making larger decisions earlier with less margin for error. A model can improve a task quickly. Rebuilding a clinical-trial network, an industrial design workflow, a tax-research practice, or a cybersecurity layer requires the company to earn trust across customers, data, people, and infrastructure.

A Portfolio Already in Motion

The launch release says Intrepid had already invested in 9 companies, including PhysicsX and StackAdapt. Its public portfolio also includes businesses across healthcare, industrial engineering, cybersecurity, tax, advertising, and enterprise software.

The portfolio mix is the practical test of the thesis. PhysicsX applies AI to engineering and physical-system simulation. StackAdapt operates in programmatic advertising. Iterative Health is rebuilding clinical-research operations around technology-enabled trial sites and AI-assisted patient identification. CoLab applies AI to mechanical design review. The products differ, but each sits close to a consequential operating workflow where accuracy, distribution, and customer trust matter as much as model capability.

Intrepid says its focus is global, with particular attention to the Canada-UK corridor and additional coverage across the United States and Europe. Toronto and London give the firm a base in 2 research-rich ecosystems while keeping it connected to larger growth markets and institutional capital pools.

What the $525M Close Changes

The final close turns Intrepid's thesis into a durable investment mandate. The firm no longer has to describe only what it intends to raise or where it hopes to invest. It can point to a fully closed inaugural vehicle, more than 80 LP relationships, 2 core offices, and 9 investments already made.

The capital does not settle whether application-layer AI companies will produce superior returns. Intrepid has not disclosed fund performance, individual check sizes, or how much of Fund I remains available. Those results will emerge through the portfolio's ability to turn technical capability into repeatable commercial systems.

For founders, the signal is clearer. Growth capital is organizing around companies that can move beyond an impressive model and take responsibility for the workflow around it. For LPs, Intrepid is offering specialist judgment about where AI value may compound after prediction itself becomes easier to buy. The $525M now has a job: follow machine intelligence into the parts of an industry where decisions, data, distribution, and trust still refuse to become commodities.