One plant flooded in September 2024 and within a week, 80% of U.S. healthcare organizations reported IV fluid shortages. Baxter International's Marion, North Carolina facility produced 60% of the nation's supply. That's the headline case. The daily reality is quieter and larger: 85 million damaged packages in 2024, up 30% year-over-year, costing businesses $4 billion.
Sean McCarthy, an early Amazon Shipping hire, spent years watching the same investigation loop repeat: query a 20-year-old warehouse system, cross-reference the carrier portal, call the driver, get no answer, fill out a 17-field claim form, wait four hours. The root problem was structural. A single shipment touches 40 to 60 processes across multiple vendors. No one funded the hundreds of integrations needed to connect them. McCarthy and Henry Ou, who led ML at Apple and built ranking systems at ByteDance, built BackOps to close that gap. Their AI agents read emails, navigate portals, call drivers, and file claims autonomously, escalating to humans only when judgment is required. Theory Ventures led their $26 million Series A.
The product's build process is worth understanding in detail: employees record screen sessions while solving problems, and BackOps converts those recordings into repeatable automated workflows via its Relay engine. The resulting numbers are 93% faster response times, 60% time savings, and 100% of eligible carrier claims filed automatically. Current customers include a top global automaker, major grocery chains, and industrial suppliers. The addressable market is $3.5 billion, growing at 13% annually. The full piece covers how Relay handles live escalation logic and why fragmentation, not technology, was always the real bottleneck.
[READ ORIGINAL →]