6Q4:How AI Is Moving from Promise to Practice

Our “6Q4” series features six questions for the leaders, innovators, and operators in the real-time analytics arena who are using innovative technologies to transform the world as we know it.

RTInsights recently asked Mayank Madan, Global Head of Data Analytics/Gen AI, at Lemongrass, what it will take for real value to start showing up as companies implement AI.

Q1: Where are CIOs and other IT leaders seeing measurable business value from AI today, and where do you think the market is still driven more by hype than outcomes?

Madan: AI isn’t new; organizations have used machine learning for industry-specific use cases for decades. What has changed is the emergence of Generative AI, large language models, and agentic AI. Today, CIOs and IT leaders are focused on using Agentic AI to boost productivity and streamline the work of teams, processes, and systems as part of everyday operations. Many organizations are running pilot projects, but only a small number have fully deployed agentic AI platforms that effectively leverage their internal processes, products, and institutional knowledge.

Q2: Given this, how do you expect AI and automation to change IT roles over the next three to five years? What skills are becoming more valuable, and which roles are most at risk of displacement or redesign?

Madan: AI and automation will play a central role in day-to-day decision making, but there will always be a human involved for critical business decisions and due diligence. AI will also be widely used as a day-to-day assistant to increase employee productivity.

Q3: As AI and analytics scale, how are IT’s data management priorities shifting, and where are organizations still underinvesting?

Madan: As AI and analytics scale, organizations will place greater emphasis on data governance, observability, and explainability. Interpreting data isn’t enough. Understanding its full lineage back to the source will be essential for explaining outcomes. As agentic AI becomes more common, especially for unstructured data, data quality and lifecycle management will be critical for building and refining high-accuracy models. Many organizations are still underinvesting in areas such as maintaining versioned datasets with complete historical, technical, and business metadata, as well as using knowledge graphs to capture data semantics and relationships. These areas should be key priorities moving forward.

For CIOs, deciding what to modernize versus maintain is always a question of business impact and ROI. With recent advances in agentic AI, modernization efforts can be accelerated to deliver acceptable returns, making it easier for the business to then align on how to reduce technical debt and choose modernizations that will enable them to use AI for their critical business processes.

Q4: Is traditional DevOps still “hot,” or is the momentum shifting toward platform engineering and developer experience? And how does that impact application performance and reliability?

Madan: There is an increased focus on developer experience and platform engineering. Instead of relying solely on pipelines and traditional CI/CD tooling, they’re shifting toward platforms that allow developers to build features more easily using self-service toolsets. DevOps remains the foundation of platform engineering, but with AI accelerating coding, testing, and documentation, the momentum is moving toward faster development enabled by more abstracted toolsets.

If done well, these platforms can improve application performance and reliability. However, overly complex or poorly managed platforms can have the opposite effect by creating performance issues and reducing reliability.

Q5: With SaaS and AI tools so accessible, is shadow IT becoming a bigger risk or a manageable reality? And how should IT governance evolve in response?

Madan: Shadow IT is becoming a greater risk for enterprises when governance and controls don’t evolve. With AI tools so easily accessible, the likelihood of IP and data leakage increases significantly. Rather than relying solely on restrictions, organizations should focus on guiding, training, and governing teams on the responsible use of AI tools. Effective IT governance should provide secure, approved alternatives within the organization’s controlled environment, reducing the need for unmanaged tools.

Q6: Finally, looking at budgets through 2026, which areas are seeing sustained increases? For example, what do you think of early-stage bets on things like quantum computing?

Madan: 2026 will see sustained growth in AI, data analytics platforms, and cybersecurity spending. As customers move their initial pilots into production, they’ll expect enterprise-grade solutions that are scalable and secure. Quantum computing remains a long-term strategic investment, with only about 5–10% of budgets typically allocated to research in this area.

About Mayank Madan: Mayank has over 22 years’ of extensive and diverse experience leading various aspects of software sales and Professional Services Business – Sales/PreSales, System Design, Solution Design, Business Intelligence, AI/ML, Deep Learning(Computer Vision AI) and Big Data Architecture and Delivery of business transformation solutions.