Artificial intelligence may be moving at breakneck speed, but the databases powering it are struggling to keep pace.
That’s the problem Weston-based engineer Harish Shenoy [pictured above] has spent much of his career trying to solve.
Shenoy, who moved to South Florida in 2018, spent seven years at Amazon Web Services, where he led the program for Amazon Aurora Limitless, a database designed to handle growing demand. Today, he’s a Data and AI Solutions Architect at Databricks, working with major retailers on the infrastructure behind AI applications.
Refresh Miami caught up with Shenoy to discuss why databases matter more than ever, what founders should be thinking about, and South Florida’s growing tech community.
Refresh Miami: Why are databases becoming such a critical issue as AI takes off?
Harish Shenoy: For years, databases were essentially the plumbing behind applications. Nobody gave them much thought until something broke. AI has changed that. Retailers want to make recommendations based on what customers are doing right now, not last week’s data. But many existing systems weren’t built to handle real-time transactions and AI-driven analysis simultaneously. That’s creating bottlenecks.
What happens when an AI startup goes from a working demo to thousands of customers?
That’s when things get interesting. The model works, the demo looks great, and then real customers and real data show up. Suddenly, you’re dealing with growing traffic, performance issues, and the risk of downtime. I’ve spent much of my career helping companies address those problems, including working on Amazon Aurora Limitless, which was designed to help databases scale as demand increases.
How is database technology changing to address these challenges?
One approach I’m particularly interested in is database branching. Developers have long been able to create branches of their code to test changes safely. Now we’re bringing that same approach to databases. You can create a copy, test a change, and catch problems before they affect customers. I’ve built open-source frameworks around this, first using AWS tools and more recently with Databricks Lakebase.
What’s one mistake companies make when managing their databases?
Assuming that having backups means they’re prepared for something to go wrong. Almost everyone backs up their data, but far fewer companies practice restoring their systems while customers are actively using them. What matters isn’t how often you save a copy. It’s how quickly you can get the business running again. Every minute of downtime can mean lost sales and lost trust.
You moved to South Florida in 2018. What attracted you to the region?
What prompted the move was the tech scene taking off here. You could see it building at the time: Amazon, Google, and Microsoft were all expanding their presence in South Florida, and startups and tech talent were following them. It was becoming a place technologists moved to on purpose, not a place they passed through.
Eight years later, how do you see South Florida’s position in the tech industry?
That growth has only accelerated. The big tech presence keeps expanding, the startup community is active, and I’ve been able to do the same caliber of work here, leading database and AI programs, while being part of a tech community that’s still growing. I see South Florida as a real place for serious technologists now.
READ MORE IN REFRESH MIAMI:
