Goldman Sachs analysts are shifting their focus from AI infrastructure giants like Nvidia to a new category of AI investment opportunities: platform stocks that enable direct AI applications. In a Thursday research note, the investment bank identified Microsoft, DataDog, MongoDB, Elastic, and Snowflake as the best-positioned companies to benefit from the next wave of generative AI investments.
The analysts argue that while Nvidia and AI infrastructure companies—including semiconductor manufacturers, cloud providers, and data center REITs—will likely continue to see share price increases, future returns may be constrained by elevated valuations. Instead, they recommend investors look toward platform stocks that provide databases and development tools, which serve as “building blocks to construct next generation applications.”
Many of these recommended platform stocks have experienced significant declines this year due to near-term fundamental weakness. However, Goldman Sachs analysts believe they now offer historically low valuations and stabilizing revisions, positioning them well as AI investment momentum rebounds. The analysts categorize AI stocks into four phases: Phase 1 includes AI infrastructure like Nvidia, Phase 2 encompasses cloud and semiconductor companies, Phase 3 consists of platforms that can monetize AI through incremental revenues, and Phase 4 represents companies that would benefit from widespread AI adoption.
According to Goldman Sachs, platform stocks represent an exception among Phase 3 investments because they can more clearly demonstrate AI monetization potential, unlike other software and IT services companies where timing remains uncertain. The analysts emphasize that “the roll-out of applications among Phase 3 stocks is a necessary condition before investors will gain confidence about owning Phase 4 stocks.”
This strategic shift comes after AI stock flows dwindled over the summer as traders expressed concerns about returns on substantial AI spending. The sector experienced sharp underperformance in July, with Nvidia tumbling as much as 27% from its June all-time high. However, the AI trade has recently reaccelerated, with Nvidia now trading near record highs again, driven by Federal Reserve interest rate cuts and strong macroeconomic data. Goldman Sachs’ recommendation suggests that the next phase of AI investment will focus less on infrastructure and more on the platforms and tools that enable practical AI applications.
Key Quotes
Our equity analysts believe ‘platform’ stocks, including databases and development tools, are set to be the primary beneficiaries of the next wave of generative AI investments. These platforms allow the best use of AI infrastructure while providing building blocks to construct next generation applications
Goldman Sachs analysts explained their rationale for recommending platform stocks over infrastructure plays in their Thursday research note, highlighting the strategic importance of companies that enable AI application development.
Expected future returns could be constrained by elevated starting valuations, although valuations are historically a poor near-term signal for large-cap equities
The analysts cautioned that while AI infrastructure stocks like Nvidia may continue rising, their high valuations could limit future returns, suggesting investors look elsewhere for better risk-reward opportunities.
We believe the roll-out of applications among Phase 3 stocks is a necessary condition before investors will gain confidence about owning Phase 4 stocks with the largest potential earnings gains from AI-related productivity
Goldman Sachs outlined their phased approach to AI investment, explaining that platform companies must successfully deploy applications before broader AI adoption can occur across the economy.
Our Take
Goldman Sachs’ pivot toward AI platform stocks reflects a maturing investment thesis that recognizes infrastructure alone doesn’t generate revenue—applications do. This recommendation is particularly astute given that companies like MongoDB and Snowflake have sold off despite being critical enablers of AI deployment. The analysts are essentially betting that the market has overreacted to near-term weakness while missing the long-term opportunity.
What’s most interesting is the implicit acknowledgment that we’re still early in the AI monetization cycle. By distinguishing between Phase 3 (platforms) and Phase 4 (widespread adoption) stocks, Goldman is telling investors to be patient and selective. The real test will be whether these platform companies can demonstrate tangible AI-driven revenue growth in coming quarters. If they succeed, this could mark an important inflection point where AI investment broadens beyond a handful of infrastructure giants into a more diverse ecosystem of beneficiaries.
Why This Matters
This Goldman Sachs analysis represents a significant evolution in AI investment strategy, signaling that the market is maturing beyond the initial infrastructure build-out phase. The shift from “picks and shovels” companies like Nvidia to platform providers indicates that investors are now seeking more direct paths to AI monetization through companies that enable application development.
The recommendation carries particular weight given the summer’s AI investment slowdown, when concerns about return on investment temporarily cooled enthusiasm for the sector. By identifying platform stocks with historically low valuations, Goldman Sachs is essentially calling the bottom on a category of AI investments that has been overlooked during the Nvidia-dominated rally.
For businesses and developers, this signals growing confidence that AI application development is entering a critical growth phase. Companies like MongoDB and Snowflake provide the essential infrastructure for building AI-powered applications, suggesting that the focus is shifting from training massive models to deploying practical AI solutions. This transition could accelerate AI adoption across industries as development tools become more accessible and standardized, ultimately bringing AI’s transformative potential closer to widespread business implementation.
Recommended Reading
For those interested in learning more about artificial intelligence, machine learning, and effective AI communication, here are some excellent resources: