Back

OSS and NMS: Getting Two Worlds to Speak the Same Language

5 MINS

OSS and NMS: Getting Two Worlds to Speak the Same Language

Most of the hard problems in telecom software aren't inside a single system, they're in the gaps between systems. The OSS wants to provision a service end-to-end. The NMS knows the actual state of the network elements. Making those two layers agree, in real time, at scale, is where I've spent a large part of my engineering career.

Interoperability is a protocol problem and a semantics problem

It's tempting to think interoperability is solved once both sides speak SNMP, NETCONF, or RESTCONF. The protocol is the easy half. The hard half is semantics: does "service active" on the OSS side mean the same thing as "active" in the NMS? Does an alarm clear because the fault is gone, or because someone acknowledged it?

Through the northbound and southbound interfaces, NBI and SBI, I've learned that every integration is really a negotiation about meaning. Get the data model right and the protocol almost disappears. Get it wrong and no amount of clever code saves you.

Service provisioning is where the money is, and the bugs

When an operator sells a service, the OSS kicks off provisioning that has to ripple down through the NMS into real configuration on real devices. Every step is a place to fail: a timeout, a half-applied config, a device that accepts the command but doesn't actually honour it.

I once worked through a TLS service provisioning flow that we optimized to over 95%, and the customer appreciation from SMART Telecom for that work stuck with me, not because of the number, but because of what it represented: a provisioning path that operators could finally trust to either fully succeed or cleanly roll back. In provisioning, partial success is the most dangerous outcome there is.

Microservices help, if you respect the boundaries

Breaking a monolithic NMS into microservices buys you scalability and independent deployment, but it also multiplies the integration surface. Suddenly the contracts between services matter as much as the contracts between OSS and NMS. I treat every service boundary the way I treat an NBI: an explicit, versioned, well-documented interface, not an accident of where the code happened to split.

The teams that succeed here are disciplined about contracts. The ones that struggle treat microservices as a way to avoid agreeing on anything.

The goal is one coherent story

When OSS and NMS truly interoperate, an operator sees one network, not two systems arguing about it. That coherence is the real deliverable, and it comes from getting the boring foundations, the models and the interfaces, exactly right.

Background

Narendra skipped presentations and built real AI products.

Narendra Billakanti was part of the April 2026 cohort at Curious PM, alongside 18 other talented participants.