GenAI initiatives fail to scale successfully due to inconsistent, outdated and biased data eroding trust and hindering decision-making. What steps should an organization take to overcome these hurdles in the fastest timeframe possible?
Garavello: Scaling GenAI initiatives successfully requires a foundation of high-quality, up-to-date, and unbiased data. Without careful attention to data integrity, AI can produce incomplete, false, or biased results—leading to ethical risks, reputational damage, and poor decision-making. Organizations must take proactive steps to overcome these challenges by ensuring data consistency, governance, and scalability
The most effective way to address these hurdles is by committing to a robust data strategy that prioritizes access to high-quality, well-harmonized, and accurate data. This involves deploying scalable data architectures with strong governance and security measures. For example, by leveraging Qlik’s data integration capabilities and utilizing the Singapore cloud region for local analytics deployment, Singlife streamlined its claims process, achieving a three-fold increase in efficiency and a 35% reduction in data analysis costs.
A key step for organizations is to enhance data quality through stringent integration and governance practices. Data from applications, transactional databases, and IoT systems should be efficiently processed into management systems such as data warehouses, lakes, or lakehouses. A federated approach enables organizations to access and utilize decentralized data without requiring full centralization, accelerating insights while reducing costs.
By synthesizing data from diverse sources, businesses can improve the robustness of AI models, reducing biases and enhancing accuracy. A well-governed data infrastructure ensures that AI solutions can deliver reliable, actionable insights—allowing organizations to scale GenAI initiatives with confidence, maximize efficiency, and drive meaningful business outcomes.
Are low-code/no-code tools the answer to these GenAI challenges? To what extent?
Garavello: Low-code/no-code tools are crucial for streamlining GenAI implementation as they enable faster development without requiring deep technical expertise. According to IDC, more than half (56%) of organizations across Asia Pacific are turning to low-code and no-code AI tools to simplify GenAI implementation. Notably, 74% see these tools as critical for AI expansion, allowing developers to focus more on higher value tasks such as creative problem solving and those with greater strategic impact.
For example, SAPPORO Holdings leveraged Qlik Cloud’s low-code development and user-friendly UI to build in-house integration with new data sources, reducing future costs for building data integration by around 80% and enabling plans which reduce development time by more than 75%.
SAPPORO can now solve issues relating to the requirements for utilizing data in business departments at an early stage and has also created a system infrastructure that will enable them to sustainably contribute to profits through such activities.
But to truly solve GenAI challenges, businesses must look at the foundational element of good AI, and that is good data that is diverse, timely, accurate, and secure. Start with AI-ready data to ensure reliability and trustworthiness in AI to set the business up for success in the long run.
