Axiom Partners' Sandhya Venkatachalam backs AI for real

Axiom Partners founder Sandhya Venkatachalam expects about half of her fund's 35 planned investments to fail. This high-risk strategy underpins her $52 million fund focused on AI for the real world, targeting industries like construction, industrials, and insurance.

Sandhya Venkatachalam's career spans building companies, product leadership at a hardware firm sold to Cisco, and a role at Skype before its Microsoft sale. She later became a general partner at Social Capital, where she led the early institutional investment in AI chipmaker Groq, and subsequently invested at Khosla Ventures. This hardware-to-software journey shaped her early interest in data and machine learning long before AI dominated venture capital. Her exploration of Google's in-house hardware efforts even led her to Jonathan Ross, a key figure in those projects who later founded Groq and championed the future importance of AI inference.

Axiom’s model targets overlooked industries with outcome-driven AI

Axiom Partners invests in AI for the real world, aiming to benefit a broader population beyond early adopters. The fund specifically targets industries underserved by technology, where AI can produce a tangible outcome rather than simply provide another software tool. This focus frequently leads to construction, industrials, or insurance. Supported companies may involve hardware, sensors, or robotics, while others are entirely software-based. The fund seeks systems that effectively deliver an outcome and steers clear of products resembling traditional enterprise tools.

Venkatachalam believes a product becomes difficult to replace when it performs significant work highly valuable to the client, a concept she describes as the 'last mile'. In practical terms, this means deep integration into client systems, understanding their data, learning key workflows, and ensuring the promised result. Clients in these sectors must be willing to pay for AI within their existing labor budgets, and Venkatachalam often sees contract values reaching hundreds of thousands of dollars.

Investment approach centers on operator partners and internal AI tool

The firm is built around active AI practitioners who work part-time, receive a share of the fund's profits, and are treated as partners. Venkatachalam argues that if you aren't building, productizing, pricing, or taking AI to market regularly, it is very hard to keep up. These operator-partners, who maintain other jobs, help keep the firm's investment thinking current. Founders are drawn to them because they have encountered similar challenges.

Axiom also uses AI internally, having built a tool called the Axiom Brain. This system helps make sense of market trends, identify interesting people and companies, and accelerate due diligence. For Venkatachalam, its core value is enabling the team to act quickly. The firm's diligence focuses on the risks that matter for a startup's next milestones, centered on two key questions: Can this team do what it says it will do? And if it can, could the result be massive?

Expects high failure rate but backs outliers in underserved sectors

While Silicon Valley often gravitates toward founders with elite backgrounds from Stanford or OpenAI, Axiom looks for more nonobvious founders in nonobvious industries. Venkatachalam accepts that many bets will not work out while aiming for the outliers that deliver major outcomes. She noted that investors are backing her to identify future categories, not to participate in ones everyone already recognizes.

"One thing I took was Vinod Khosla’s open view of where great founders can come from," said Sandhya Venkatachalam. Axiom Partners plans to make 35 investments in AI startups focused on construction, industrials, and insurance, with an expectation that about half will fail.