.png){kind=logo}
AI investments have surged globally, and a stark reality now confronts enterprises. 85% of AI projects fail not for their algorithmic weaknesses but because of poor data quality and infrastructure.
In a recent episode of the Analytics Insight podcast, host Priya Dialani engaged in an insightful conversation with Arun Reddy, Senior Cloud Database Engineer at CoStar Reality Inc. The discussion centered on how data forms the foundation of AI innovation and why robust data strategies are critical to the success of generative AI projects.
CoStar Reality has a presence in the USA, UK, and Australia, and it provides its customers with accurate property information through data on sites like Apartments.com and Homes.com. These platforms make it easier to access property listings, rentals, & even insights on commercial real estate.
Arun talked about his role in ensuring data accuracy, database performance, and smooth data flow across both cloud and local servers. His contribution helps AI models analyze trends, improve customer service, and provide quicker insights.
Arun is skilled at optimizing data pipelines, reducing processing time from 96 to 6 hours, and backing up CoStar’s data systems during high-demand periods like the Super Bowl.
He is a versatile and skilled data engineer who is dedicated to automation, flexible in scaling, and eager to learn new technologies. These traits make him ideally suited to the evolving needs of AI.
Arun emphasized that the factors of quality, governance, and security in AI ecosystems cannot be ignored. The use of data validation, encryption, and adherence to standards like GDPR & HIPAA are the basis of trust and reliability.
He pointed out that using AWS KMS, SQL Server Encryption, and Azure Purview is indispensable for protecting sensitive data. Security integration within every stage of data collection, processing, and storage becomes necessary as AI turns more complex.
Looking ahead, Arun Reddy predicted a shift toward real-time data processing, hybrid search models, and smaller domain-specific AI systems.
He further thinks that the organizations ready for the future will not only adopt privacy-by-design frameworks but also reap the benefits of automation, allowing them to be quick in response. “AI is only as good as the data it learns from,” he said, reiterating that a strong data foundation will be the hallmark of the next generation of AI innovations.
