Not just Big Tech: Which small-cap companies are benefiting from the development of AI
The excitement around artificial intelligence is heating up not only the securities of technology giants. Against the background of record quarterly reports of Big Tech and growing optimism of Wall Street, investors are finding undervalued ideas in the small capitalization segment. The biggest gainers are companies creating infrastructure for AI development - from building data centers to supplying specialized equipment and solutions.
Analyst Aldiyar Anuarbekov selected 46 companies from Russell 2000 index for Oninvest, whose main specialization is related to AI technologies, and calculated an equal weighted index, named AI Small Cap EW (equal weight) Index. It includes manufacturers of data center equipment, developers of sensors and computer vision systems, robotics and quantum computing solutions.
Since the end of November 2022, which can be considered the starting point of the AI boom after the launch of ChatGPT, the AI Small Cap EW Index has gained 174.5%. By comparison, over the same period, the Nasdaq-100, which includes the largest U.S. technology companies, has added 98.7%, the S&P 500 has added 64.4%, the SOXX index of semiconductor manufacturers has gained 98%, and shares of Nvidia - the main beneficiary of AI - have seen explosive gains, up 975.6%. Among the top gainers in the AI Small Cap EW Index are quantum technology developers, optical component suppliers and companies that build infrastructure for data processing.
Growth leaders and outsiders
Another, conditionally second starting point for the active phase of the AI boom is mid-2023. It was then that a turning point began to emerge: the government accelerated steps in regulation - in particular, leading players assumed voluntary obligations to manage AI-related risks. At the same time, businesses began to confidently implement generative AI: according to McKinsey, one-third of companies surveyed were already actively using it. Since mid-2023, of the 46 companies in the AI Small Cap EW Index, more than half - 24 - have grown, while the rest have declined.
In 2025, the situation has changed: since the beginning of the year, only 19 companies remained in the plus side, while 27 were in the minus side. In general, the dynamics of the index is mainly supported by several strong companies from the infrastructure, defense and niche AI services sectors. At the same time, the dynamics within the segment varies greatly. The greatest growth was shown by the areas related to infrastructure for AI - optical modules, data centers, network equipment - as well as quantum computing, robotics and specialized AI services, including voice interfaces and analytical platforms. Here, growth was supported by real demand and increasing order volumes.
At the same time, some niches remained on the sidelines: AI biotechnology, additive manufacturing, individual AI platforms without a steady flow of customers, as well as some companies from the insurtech (technologies for the insurance sector) and adtech (technologies for the advertising industry) segments failed to demonstrate comparable growth. Probable reasons were long commercialization cycles, market saturation and high competition, which limited growth.
Aldiyar Anuarbekov highlighted three companies - leaders of growth from the beginning of 2025.
Pagaya Technologies
Pagaya Technologies is a fintech platform in consumer lending, its AI-powered infrastructure enables banks, credit unions and BNPL services to approve more loans without compromising portfolio quality. Pagaya's AI model analyzes huge data sets and finds solvent customers where traditional scoring systems fail. Pagaya is already processing millions of applications per year, and connecting each new partner reinforces the value of the entire ecosystem - the network effect becomes visible. The business model is scalable due to the diversification of the product line: from student and car loans to mortgage programs, which creates additional growth drivers for the company, UBS noted in its review.
Pagaya delivered an impressive financial breakthrough in the second quarter of 2025, with network volume up 14% year-on-year to $2.6 billion - more than the company had forecast. Revenues increased by 30% to $326 million, while adjusted EBITDA grew by 72% to $86 million. It is especially noteworthy that GAAP net income amounted to $17 million, while a year ago there was a loss of $75 million.
Since the end of November 2022, Pagaya's stock is up 187.3%. Since the beginning of 2025, the company's stock price has soared about 270%, while the Nasdaq-100 is up 12% and the S&P 500 is up 10%. Following the release of the quarterly report, analysts at Jefferies reiterated a Buy rating and $46 target for Pagaya stock. The new price target implies the growth of the securities' value by about 30% from the current value.
Jefferies emphasizes that Pagaya's business model has a strong network effect. UBS analysts noted in their report that in an environment of high interest rates and tightening credit standards, Pagaya's key advantage remains the ability of AI algorithms to find solvent borrowers among the customers most likely to be rejected by traditional banks. About 70% of the applications that Pagaya receives have been rejected by banks, while the probability of defaults remains within market averages, said Pagaya CEO Gal Krubiner.
Aeva Technologies
Aeva Technologies specializes in next-generation lidar technologies (Light Detection and Ranging). The company manufactures FMCW lidars that can use a laser to simultaneously measure the distance and speed of objects, providing accurate data and resistance to interference, even in difficult weather conditions. AI algorithms enable lidars to recognize objects and predict their trajectories. McKinsey estimates that the market for automotive AD components (for autonomous driving systems), which include lidars, cameras and radars, could grow to $55 billion to $80 billion by 2030. In addition to unmanned trucks and cars, the company's technologies are used in Tampa International Airport (Florida), D2 Traffic control systems, critical infrastructure in collaboration with Sandia National Labs for nuclear power plants, and in aviation - in the Airbus UpNext project.
In the second quarter of 2025, Aeva Technologies' revenue grew sharply - up 175% year-over-year to $5.5 million - driven by increased demand for its 4D LiDAR technology and expanded partnerships. At the same time, the loss widened significantly to $192.7 million from $43.4 million a year earlier, reflecting the company's significant research and development (R&D) and business scaling costs. Nevertheless, the company reaffirmed its revenue outlook for 2025: it now forecasts 100-110% growth, up from 80-100% previously.
In the trucking segment, Aeva Technologies works with Daimler Truck and Torc. Among the partners is LG Innotek, a subsidiary of LG Group, which produces electronic and optical components. It has already invested $77.5 million in Aeva's capital, emphasizing its long-term focus on 4D LiDAR development.
Aeva shares are up 78.5% since the end of November 2022 and nearly 200% since the beginning of 2025. Risks remain, from the need for new capital to reaching profitability. But the rare combination of large customers, technology headroom and a growing market for small-cap makes Aeva one of the standout stories in the lidar segment. Following the release of its second-quarter report, Canaccord Genuity maintained a Buy rating on the company's stock, increasing the target from $16 to $24, implying a potential upside of 60% from current value.
Applied Digital
Applied Digital builds and manages data centers focused on generative AI, machine learning and high-performance computing. Since the beginning of 2023, Applied Digital's stock has gained 266%, up 9.8% in 2024, and already up 111% since the beginning of 2025. The growth was driven by contracts with CoreWeave and other customers to host capacity for AI models, as well as expansion of data centers in key U.S. regions.
For the fourth quarter of fiscal 2025, Applied Digital's revenue grew 41% year-over-year to $38 million. The company posted a small positive operating profit on a "net" adjusted EBITDA basis - EBITDA was $1 million - and narrowed its adjusted loss to $7.6 million. For the full year, revenue added 6% year-over-year to $144.2 million, and EBITDA reached $19.6 million. Following the report, analysts at Citizens JMP maintained their Outperform rating (buy recommendation), increasing the target price from $12 to $18. The new target implies the shares growth by almost 10% more.
Applied Digital's key pillar is long-term contracts that produce predictable cash flow. In early June, the company entered into a 250 MW capacity lease with CoreWeave for about 15 years, which could generate about $7 billion in revenue. After Applied Digital published its quarterly report, the companies agreed to add another 150 MW - the final contract grew to 400 MW and about $11 billion.
At the same time, the company continues to invest in new data centers and energy efficiency technologies. Goldman Sachs forecasts that the global data center market will grow from 55 GW to 84 GW by 2027, with AI workloads accounting for 27% of the market. Applied Digital is among the key beneficiaries of the growing demand for high-performance AI centers. However, the high capital intensity of the segment and the need to simultaneously scale the business and control costs remain major challenges for the company.
This article was AI-translated and verified by a human editor
