How Greenlight Scales FinTech SRE Operations with AI
Greenlight CIO and CSO Prabhath Karanth on running FinTech infrastructure with agentic workflows and aiming for pin-drop silence in production. Watch the talk.
Greenlight is a family finance platform that moves money, runs payments, and supports safety and hardware products, on infrastructure with thousands of microservices and hundreds of third-party connections. At FLOCK '26, Prabhath Karanth (PK), CIO and CSO at Greenlight, talks with Andrea Sy, Director of Product Marketing at NeuBird, about keeping that platform reliable as it grows.
PK's approach is to turn every repeat task into an agentic workflow. Incidents follow common patterns, so a recurring one should take half the time to solve the next time, then a quarter, and eventually stop happening. He measures success by results: incident volume low enough to reach what his CEO calls pin-drop silence. Coming from the security industry, he sees NeuBird bringing overdue innovation to operations, and compares the goal to an MRI of the whole infrastructure in real time that finds issues before they turn into outages.
For teams starting out, his advice is to pick one or two priorities, such as cost, an agentic SRE workflow or early warning of infrastructure problems, and double down before expanding. In audience questions he explains why he runs security incident response and SRE on one 24/7 on-call rotation, since the same logs serve both, and why defending at machine speed is now the only option.
In this conversation
Turning repeat incidents into agentic workflows
Keeping a premium experience without breaking the bank
What success looks like: pin-drop silence
Why operations is ready for the innovation security has seen
Advice: pick one or two priorities, then double down
Running security incident response and SRE on one on-call
Related Resources
See NeuBird in action
Root cause in minutes, not war rooms.
