Episode 59: JetStor’s Jim Gallagher on Why 'Good Enough' Infrastructure Costs More Than You Think

Download, Listen, and Subscribe

Apple  |  Spotify  |  YouTube

Or search “Manufacturing the Future” wherever you listen to podcasts!

"Every company is becoming both a data company and a bank. If you're not doing this stuff and not keeping an eye on it, other people are."

A lasting warning from Jim Gallagher, CEO at JetStor, who argues that although manufacturers are generating more data than ever, most haven't built the infrastructure to actually use it. The common mistake isn't a lack of storage; it's treating storage as a flat, one-size-fits-all decision rather than a tiered architecture matched to how data gets used. Without that foundation, companies can't run real-time analytics, can't prepare for AI workloads, and are quietly accumulating technical debt that compounds over time.

Jetstor is a US-based enterprise storage company with more than 30 years in the business. The company designs and manufactures scalable storage systems built for demanding workloads, including virtualization, high-performance computing, media production, and data-intensive manufacturing environments.

In This Episode:

Jim walks through how manufacturers should be structuring their data strategy, starting with a three-tier classification framework: tier 1: for real-time, latency-sensitive workloads; tier 2: active archive for data that still needs to be accessible; and tier 3: deep archive for long-term retention. He explains why staying with legacy infrastructure isn't actually "free.” Jim closes with a concrete challenge for manufacturing leaders: when was the last time your team actually tested your backups?

Topics

• Why storage is not a flat ecosystem, and the performance-cost trade-offs that actually matter
• Three-tier data classification: real-time, active archive, and deep archive
• The "data lake" trap: why unstructured data hoarding happens and what it actually costs Why training workloads and inference workloads need entirely different architectures
• The 1-2% annual hardware failure rate and what that means for legacy infrastructure planning
• How the DevOps movement in IT foreshadows the IT/OT convergence coming to manufacturing 
• Why "when did you last test your backups" is the question manufacturing leaders should be asking right now
• Ransomware as a business risk, data insurance products, and what underwriting requirements actually look like
• Why manufacturers that have been gathering data for decades may be sitting on unexpected revenue streams