AI Beyond Platforms - How Data Will Unlock New Value In 2025

In 2025, the artificial intelligence (AI) landscape is undergoing a transformative shift. No longer confined to the technology infrastructure layer—where foundational models and hardware dominate—AI is expanding into two critical new layers: the application layer and the data infrastructure layer. These layers are poised to redefine how businesses operate and make decisions, with the potential to unlock unprecedented value.

Venture funding for AI companies surged to $99.6 billion in 2024, an 80% increase from the previous year, according to Crunchbase data. While nearly a third of this funding targeted foundation model companies, the remaining two-thirds flowed into sectors heavily impacted by these models, such as autonomous driving, healthcare, robotics, professional services, and marketing.

This massive investment underscores the growing emphasis on the practical application of AI. Companies and investors are increasingly focusing on how AI can be harnessed to create tangible outcomes, signaling a significant departure from the platform-driven phase of the past.

Shifting to Applied AI

In 2023, much of the focus was on the tools and infrastructure required to build and maintain AI models. By 2024, attention shifted toward applied AI, with a notable emphasis on developing applications that directly impact specific industries. Investors have recognized the importance of identifying solutions that meet enterprise needs effectively, balancing the probabilistic nature of AI outputs with practical use cases. Success in this area often involves determining when AI’s performance is sufficient for real-world applications.

The Application Layer: Transforming Business Processes

The application layer represents AI’s integration into specific workflows and industries. Pre-built solutions are becoming more popular as enterprises seek ready-to-use applications that simplify adoption and deployment. Reports highlight that organizations are still navigating this new terrain, learning how to implement AI effectively to achieve desired accuracy through iterative processes. Pre-built applications provide a more accessible entry point for businesses.

Generative AI is particularly promising, transforming service-based revenue models into scalable product offerings. Companies like Unstructured exemplify how generative models streamline data processing and transformation tasks. These applications are revolutionizing industries—from marketing to professional services—by providing more efficient and automated solutions.

The Data Infrastructure Layer: Fueling AI’s Potential

AI’s efficacy relies heavily on the quality of the data it consumes. The data infrastructure layer encompasses the tools, platforms, and suppliers that provide, process, and manage this data. High-quality, well-structured data is essential for training accurate and reliable AI models.

The Role of Quality Data Providers

Suppliers of zero-party data play a pivotal role in the data infrastructure layer. This data, collected directly from consumers through surveys and other consent-based methods, provides reliable insights into spending intentions, shopping habits, and economic attitudes. Over decades, zero-party data has enabled predictive analytics for retail stock revenues and macroeconomic indicators, offering actionable insights for various industries.

Similarly, companies like LSEG specialize in financial and alternative data, enhancing investment strategies and economic forecasting. Providers like this bridge the gap between raw data and actionable insights, ensuring that AI applications can deliver meaningful outcomes.

Ereteam, a company renowned for its advanced predictive analytics, also plays a significant role in this space. Its collaboration with other data suppliers has produced publicly validated retail stock revenue forecasts with over 95% accuracy, demonstrating the transformative potential of integrating high-quality data with cutting-edge analytics. “The effectiveness of predictive analytics powered by AI and Machine Learning ultimately hinges on the quality of the training data—far more than on the choice of any particular model,” explains Dr. Demirhan Yenigun, Chief Strategy Officer at Ereteam.

A research team at Exponential (Exponential-Technology.ai), have further exemplified innovation in this field by developing a highly accurate forecast of the global macroeconomic numbers such as the Consumer Price Index (CPI). Using AI and higher quality data sources, this forecast is available weeks before the government’s official release, showcasing the power of AI in delivering early, actionable economic insights that drive informed decision-making. Exponential also develops an AI-first data infrastructure to accelerate agentic AI analysis of data in the investment industry. Exponential’s Founder, Morgan Slade, says, “We now have software (Gen-AI) writing software thereby reducing the scarcity of software so high-quality data has become the new scarce good, the new differentiator.”

Agentic AI: The Next Frontier

A burgeoning area within the application layer is agentic AI, which moves beyond isolated tasks to owning end-to-end workflows. This capability enables AI to handle complex jobs autonomously, from data analysis to decision-making. However, creating effective agentic AI solutions remains a challenge due to the intricacy of integrating various processes seamlessly.

Industry Applications: A World of Possibilities

AI applications are reshaping industries by driving efficiency and innovation:

Retail and Consumer Insights

• Predictive models leveraging zero-party data enable businesses to anticipate consumer behavior and adjust strategies accordingly.
• Data-driven insights into shopping habits and motivations help retailers optimize inventory, marketing, and customer engagement.

Financial Services

• Hedge funds and investment platforms use alternative data from providers like Prosper Insights & Analytics to generate alpha and forecast market trends.
• Predictive analytics from zero-party data sources aid in stock recommendations and macroeconomic forecasting.

Healthcare

• Specialized data suppliers like IQVIA provide healthcare data that powers AI-driven diagnostics, personalized medicine, and operational efficiencies. This momentum is now bolstered by the White House’s newly announced Stargate initiative—an AI infrastructure investment of up to 500 billion dollars focused on healthcare. During the press conference, Ellison highlighted AI’s ability to transform health care, saying that the tools provided by OpenAI and Softbank are helping to develop a cancer vaccine.
• “You can do early cancer detection with a blood test, and, using AI to look at the blood test, you can find the cancer that is seriously threatening the person,” Oracle’s Larry Ellison said. “Once we gene-sequence that gene tumor, you can vaccinate the person against that cancer, and you can make that mRNA vaccine robotically using AI within 48 hours.”

Marketing and Sales

• AI applications are transforming lead generation, customer segmentation, and campaign optimization, making marketing efforts more targeted and effective.

Economic Forecasting

• Exponential’s CPI forecasting model demonstrates how AI can deliver early economic indicators, providing businesses and policymakers with crucial data well in advance of official reports.

The Importance of Data Quality

As AI moves deeper into the application and data infrastructure layers, the importance of data quality cannot be overstated. Accurate, timely, and relevant data ensures the reliability of AI models and their outputs. Suppliers that prioritize compliance with regulations such as GDPR and HIPAA, while maintaining high data standards, will be indispensable to the AI ecosystem. This is why Exponential says their Unifier data warehouse, and new data infrastructure like it, will be crucial to building agentic AI solutions at scale. Orchestrating and maintaining massive high-quality datasets in real-time, at scale is one of the most complex steps in the entire process.

Investment Trends and Future Outlook

The substantial investment flowing into AI’s new layers reflects their potential to revolutionize business operations. The focus on applications and data infrastructure signals a maturation of the AI industry, where the emphasis is on solving real-world problems and generating measurable outcomes.

Companies that can successfully integrate AI into workflows—fueled by high-quality data—are poised to lead this new era of innovation. The transition from traditional SaaS dominance to AI-driven solutions marks a significant evolution in technology.

Conclusion

In 2025, AI is no longer just about building platforms—it’s about applying data to generate transformative outcomes. The application and data infrastructure layers represent the future of AI, offering businesses the tools to innovate, optimize, and grow. With significant investments and groundbreaking advancements, these layers are shaping a new horizon for AI—one where technology moves from theoretical potential to practical impact, redefining how companies operate and make decisions.

The journey from platforms to applications is a pivotal moment for AI. As enterprises embrace this shift, the focus will increasingly be on delivering results that are not only innovative but also actionable and impactful, heralding a new era of AI-driven success.