As we move forward with AI adoption and integration, the relationship between AI and data becomes more symbiotic and pivotal for innovation. We explore two emerging focuses that are set to shape the future of data analytics and AI in 2025, according to Lingaro experts.
For years, enterprises have latched on to the promise of game-changing business insights from massive datasets. However, Lingaro’s experts observe a big shift in focus that’s emerging in 2025.
The year 2025 is poised to be transformative for data analytics and artificial intelligence (AI). “The more data, the merrier” mentality is coming to a sunset, and more diverse and focused datasets are now in the spotlight. In the 2024 Hype Cycle for Data, Analytics and AI Programs and Practices report, Gartner shares that 70% of organizations will transition from big data to wide data by 2025.
Because of this big shift, two pivotal focuses in data analytics emerge this year: "AI for Data" and "Data for AI." These focuses highlight the symbiotic relationship between AI and data, where AI enhances data management and analytics, and high-quality data fuels AI capabilities.
Figure 1. The symbiotic relationship between AI for Data and Data for AI
AI for Data: Enhancing Data Processes
Sammilan Dey, Head of Lingaro’s Data Consulting Practice, says that as AI continues to mature, enterprises become more equipped with advanced tools that will ease data processing tasks. Due to this trajectory, focusing on “AI for Data” becomes paramount for businesses to stay competitive.
“AI for Data” refers to utilizing AI technologies to enhance data-related processes. Focusing on this approach enables enterprises to reallocate efforts and resources toward AI adoption and integration for a more agile and scalable business.
Increasing GenAI Integrations
One of the most notable examples of “AI for Data” is the increasing integration of generative AI (GenAI) into existing data ecosystems. The outlook on GenAI has transformed in the recent past, according to Harish Ravi, Head of Data Consumption in Lingaro.
GenAI has shifted from being seen as a specific component to an optimized, additional layer across various data assets and tools. For instance, Copilot and Einstein Analytics are now utilized as chatbots integrated to tools such as Power BI and Tableau. Chatbot integrations like this can provide users with more insights from existing data models and reports without the need for additional data processing steps.
GenAI integrations transform how enterprises can leverage their data assets — automating manual work, enhancing data quality, and extracting insights with speed. This can lead to ideal outcomes such as faster data processing, better structured data, and more accurate insights for better decision-making.
It also impacts organizations from a cost efficiency perspective. Gartner predicts that as we reach 2027, over 40% of digital workplace operations will use GenAI-enhanced tools, significantly lowering resource requirements.
GenAI integrations with existing data assets no longer need specific data architectures for GenAI, which are both labor- and cost-intensive. Based on McKinsey’s estimates, building a GenAI foundation model for a stand-alone business case costs around US$5 – US$200 million. This includes model development, data and model pipelines, model training, and plug-in-layer building. Meanwhile, GenAI model integrations with internal data and systems only cost around US$2 – US$10 million, including only fine-tuning the model and building data and model pipelines, as well as plug-in layers.
Supporting a Multinational CPG Corporation Pursue AI for Data
At Lingaro, we support enterprises to pursue “AI for Data” through innovative GenAI integrations. For instance, we partnered with a multinational consumer goods corporation to find a solution for their challenges in querying data using natural language and generating summaries for market reports.
To address their issues and enhance their market reporting process, we delivered a GenAI solution that can access their data in tables, make calculations, and create summaries from simple user questions. The solution also automated their market summaries, which led to report delivery four times faster than manual SQL and Power BI dashboard analysis. Overall, the organization increased their productivity and accuracy in creating detailed business reports with the solution.
Data for AI: Demand for High-Quality Data
AI algorithms are often "off-the-shelf" with standardized solutions that can be easily adopted by enterprises. But according to Dey, the true differentiator in achieving insightful AI outcomes is the quality of data used.
“Data for AI” highlights the need for high-quality and well-structured data. Focusing on this approach enables businesses to prioritize data quality to realize the ROI on AI pursuits.
The Necessity of High-Quality, Well-Structured Data
High-quality and well-structured data is crucial for enterprises that aim to successfully leverage AI. Gartner also emphasizes that realizing the full benefits of AI is impossible without improving data quality (Hype Cycle for Data, Analytics and AI Programs and Practices, 2024).
In line with this necessity, chief technology officers (CTOs) and chief data officers (CDOs) must stay on top of maintaining the integrity and reliability of data. This involves implementing robust data quality assessments, data cleaning processes, and data governance practices. In doing so, enterprises can have data that is suitable for AI model training and at the same time, compliant with regulations.
Integrating diverse data sources and data management are also critical for enterprises to ensure data consistency. Data integration from databases, cloud storage, and external APIs to create a unified dataset helps provide a comprehensive view of data for AI model training. Automating data pipelines to ensure that data is processed and transformed correctly, as well as effective data management practices, contribute to up-to-date and readily available data for AI.
Helping a Multinational FMCG Company Achieve Data for AI
As dedicated data partners, we understand the criticality of helping organizations build solid foundations to realize their data’s full value. That is why we are keen on giving enterprises full support in ensuring the consistency, accuracy, and reliability of their data.
As a recent example, we worked with a multinational fast-moving consumer goods (FMCG) corporation that confronted issues in data quality visibility and standard reporting. To address their problems, we delivered a data quality framework with a “No-Code UI,” which allowed users to view and manage data quality rules without technical expertise. Additionally, we included a mechanism to receive data quality issues detected in upstream systems, which enabled faster data quality checks for newly added assets.
The Road Ahead with AI and Data
The symbiotic relationship between "AI for Data" and "Data for AI" is reshaping the landscape of data analytics and AI. Because organizations are pivoting from data quantity to data quality, this relationship is more than just a trend. It’s becoming a fundamental shift that will redefine how enterprises operate this year.
By harnessing AI to enhance data processing and analysis, organizations can streamline operations and gain access to actionable insights and well-informed decisions with less time and effort. Conversely, high-quality data is crucial for training and refining AI tools and models to make sure they yield precise and dependable results.
As we move further in 2025, the integration of AI and data will become even more seamless, driving innovation and efficiency across industries. Organizations that invest in AI and data together will be better positioned to seize the full potential of their data and AI capabilities, while remaining competitive in the evolving digital landscape.
