6 months ago
38
Data and AI are two sides of the same coin in the digital era. However, what often goes unnoticed is data’s role in enabling AI’s potential.
At Cypher 2024, India’s Biggest AI Conference by AIM Media House, Pradeep Gulipalli, Co-Founder and CEO of Tiger Analytics, India, said, “AI, at its core, thrives on data”.
But data doesn’t just land in our laps, ready to be analysed. It’s a raw, often unstructured resource that needs to undergo several processes before it becomes valuable.
“This sentiment resonates with many in the industry who understand that a crucial preparatory phase ensures data is usable before data scientists can apply machine learning models. Without this foundation, even the most advanced AI systems would struggle to deliver meaningful outcomes,” Gulipalli said.
Gulipalli’s take on the early steps of data processing is straightforward: “We’re all familiar with how important AI models are, but very few appreciate what happens before that. The data ingestion, cleaning, and harmonisation often require as much attention as the modelling phase itself.”
Pradeep Gulipalli, Co-Founder and CEO of Tiger Analytics, IndiaThe complexity of the modern business environment means that data is generated at every step of an organisation’s journey. He noted that any company that manufactures cars or offers financial services produces vast amounts of data at each stage of its operations. Whether sourcing raw materials or collecting financial transactions, the data comes from multiple sources and in different formats.
This becomes even more evident when considering large organisations with diverse functions, such as supply chain management, marketing, customer service, and R&D, which all generate unique data sets.
“The reality is that we have no shortage of data. We’re often missing a unified system that can make sense of it,” he said.
How to Unlock the Potential of Data?
For Gulipalli, “data products” are the key to unlocking the full potential of data. He explains: “In a world where enterprises have hundreds of data sources, it’s no longer about wrangling each data set individually. Instead, we need to consider domain data products—consolidated, clean data serving specific business functions.”
This shift in thinking is transforming how organisations manage data. In Gulipalli’s view, “AI plays a crucial role in analytics and the entire data lifecycle. It helps with data ingestion, automates cleaning processes, performs harmonisation across sources, and even creates domain-specific models.”
For example, AI’s role in transforming unstructured data into structured formats has significantly reduced the time and effort traditionally needed to prepare data. According to Gulipalli, “We’re at a point where AI can handle a lot of the grunt work that used to be manual. It can look at data quality, detect anomalies, suggest corrections, and even propose optimal transformations—all in real-time.”
This automation allows businesses to act on insights faster, leading to better decision-making. A data scientist might focus on customer segmentation, but thanks to AI, new patterns or insights from unexpected areas within the organisation might come to light. “It’s like having an extra layer of intelligence that’s constantly improving the quality and utility of the data,” Gulipalli emphasises.
On the other end, if organisations require data from different sources, there is a need to make it affordable to access.
Shekar Sivasubramanian, CEO of Wadhwani AI, mentioned, “Collecting data like X-rays and MRIs required to work on health-related innovations in AI is expensive. An X-ray copy costs INR 100, but taking a photograph of the X-ray is free.”
Interpreting an X-ray as a photograph might seem unconventional, but it’s a legal and practical approach that reduces costs. So, collecting data as an affordable resource for AI innovation is crucial.
AI’s Ability to Impact Data Operations at Scale
“Historically, data has been treated as static—a process that involves several steps, often executed separately. But with AI, we’re shifting towards an integrated view where data and AI work together from the beginning of the data pipeline.” This integrated approach doesn’t just improve efficiency; it also enhances the accuracy of data-driven decision-making.
The benefits of adopting this AI-first approach to data management are becoming apparent. He pointed out that businesses now realise that AI is not just about analytics. It’s about automating and optimising every phase of data handling, from ingestion to cleaning to model building. It streamlines operations and helps unlock the actual value of data.
Overall, it highlights a broader industry trend: the growing recognition that AI is no longer just a tool for advanced analytics but is becoming the foundation for modern data management.
The future lies in treating AI and data as interconnected systems. When organisations embrace this AI-first mindset, they’ll unlock previously unimaginable levels of efficiency and insight.
