Take back your telco
The core problem: Modernizing your telco forces a bet-the-farm choice between vendors—and your career. But now, with AI capabilities evolving at breakneck speed, three- to five-year vendor migrations optimize for a world that won’t exist when you finish. Luckily, there’s a better way: owning your own context layer. When you build an ontology for your telco, you get a semantic model that creates options across your existing systems instead of locking you into any vendor’s roadmap.
True story: a major operator’s CIO recently told us they want to migrate off Amdocs. There’s just one problem. Amdocs has done so much code-level customization over the years that the operator doesn’t actually know how its own offers work. The knowledge lives in Amdocs consultants’ heads. Not in documentation. Not in a model. In the brains of people it rents by the hour. And obviously, Amdocs isn’t going to make leaving easy.
That’s not a technology problem. That’s a hostage situation.
And this operator isn’t alone. For the last two decades, every serious attempt to escape legacy has looked the same: switch core vendors, consolidate onto a single suite, or launch a multi-year “modernization” program. If it works, you get a newer version of the same architecture. If it doesn’t, you’ve burned years, hundreds of millions, and political capital.
Underneath it all is one simple problem: you don’t own the context layer of your own business. Now, there’s a way to stop playing the vendors’ game. It starts with owning the one layer no vendor wants you to control.
How you currently manage your context
When people say “our systems are complex,” what they usually mean is “our context is undocumented.”
Ask your team or your vendor basic questions like:
- “What has to change when we launch this new offer?”
- “Which downstream systems depend on this field?”
- “What breaks if we onboard this partner differently?”
The answer is almost never “Here, let me show you the model.”
The answer is a meeting. Someone summons the right architect, the right vendor expert, and the right consultant so they can reconstruct reality from memory, tribal knowledge, and a few out‑of‑date diagrams. You know what I’m talking about.
That’s survivable when you change one process a quarter. But when you want your business to be agile and move quickly across marketing, care, networks, and operations, it’s a dead end. You are not in control of your own destiny.
Even worse, your predicament is not an accident. It’s by design. Amdocs generates the majority of its multi‑billion‑dollar revenue from headcount‑dependent services—with billions annually in managed services alone. Those consultants perform the semantic translation between systems that Amdocs itself built without interoperability. Pretty much every other vendor in telco runs the same playbook.
You have to spend tens of billions on professional services because your context lives in their heads, and they bill you to access it. The vendors who built this mess won’t sell you a way out. Doing so would eliminate the consulting layer their business depends on.
Your get-out-of-jail card
So, how do you escape? Well, thankfully AI came around, and it’s going to save you.
Start by separating your context from your systems. Instead of treating each system as the source of truth for its slice of reality, you pull that reality out into a separate layer. You create an independent, shared model of how your telco actually works: products, offers, subscribers, balances, usage, network assets, processes, events.
And here’s the thing: that shared model is precisely what AI needs. Not more data. Not smarter models. A consistent, correct view of your business that it can actually act on.
That model is your context layer—a telco ontology. An ontology is a graph of concepts and relationships that describes your business independently of how any one vendor implements it. It’s the difference between having to ask, “What does System X think a subscriber is?” and being able to say, “Here’s what we mean by a subscriber, and here’s how System X maps into that.”
Once you own that layer, two things change immediately:
- AI plugs into a consistent, telco‑specific view of the world instead of having to reverse‑engineer semantics from every system. New use cases reuse the same context instead of redefining it.
- Modernization stops being a clean‑room rewrite and becomes a series of controlled, reversible steps.
If you free your context, suddenly you have tons of options. And if you free your context, you’ll be ready for AI.
Why options matter now
AI capabilities are evolving on Huang’s‑Law timescales: doubling every year or faster. Vendor migration programs run on telco timescales: three to five years if you’re lucky. By the time you finish consolidating around Vendor A, the AI landscape will have shifted so dramatically that Vendor A’s architecture may already be obsolete. Don’t sink yourself into a consolidation program that optimizes for a world that won’t exist when you get there.
An ontology completely inverts this dynamic, allowing you to stop betting on which vendor will win and start making your architecture indifferent to the answer.
You can move a product catalog or charging function without re‑explaining your entire business to every downstream system. You can trial a new AI use case in one market, using your ontology, without rebuilding context. You can tell any vendor: “This is our model. You map into it, not the other way around.”
The question shifts from “Do we bet the farm?” to “What’s the next small, reversible bet we can make on top of the context layer we already own?”
And there’s a compounding effect: every action that flows through the ontology enriches it. Edge cases reveal gaps in definitions. Success and failure data improves decision logic. Invalid attempts expose missing business rules.
The system gets smarter as you use it, giving you a living system that compounds with every decision.
Take back your telco
This is what the Totogi Ontology gives you: a telco‑specific context layer, and a shared semantic model that sits above your existing estate and becomes the brain your systems plug into.
You don’t rip anything out to start. Begin with a narrow, painful slice—dormant cells, change requests, offer changes—map it into the ontology, and let AI work against that clean model. As you expand, more of your business runs on context you own instead of context you rent.
Remember that operator trying to escape Amdocs? The way out is to build an ontology. Use the migration to capture its own context, break the dependency, and make sure no vendor ever holds its business hostage again. That’s what the Totogi Ontology is built for. Your systems don’t need to change. You need to own your own context.
Recent Posts
Get my FREE insider newsletter, delivered every two weeks, with curated content to help telco execs across the globe move to the public cloud.
Get started
Contact Totogi today to start your cloud and AI journey and achieve up to 80% lower TCO and 20% higher ARPU.
Discover
Before you invest in AI solutions, ask vendors this: Can you show me the money? Most can’t. But Totogi can. Watch DR’s talk to find out why.
Engage
Set up a meeting to learn how the Totogi platform can unify and supercharge your legacy systems, kickstarting your AI-first transition.
Dive Deep
The Totogi Ontology sits atop your existing BSS, OSS, and network, providing the business context that enables AI to work across your entire estate.
Frequently Asked Questions
It’s not an accident. It’s a business model. Vendors like Amdocs generate billions annually from managed services, which means their consultants become the living documentation of how your systems work. Every customization they build deepens the dependency. Every year that goes by, more institutional knowledge lives in their heads and on their billable timesheets. When you work with them, you’re not buying software. You’re renting the ability to understand your own business. The path to freedom is to stop letting any vendor hold your context hostage in the first place.
A telco ontology is a semantic model—which is a graph of concepts and relationships that describes how your business actually works, independently of any vendor’s implementation. Think of it as the difference between asking “What does Amdocs think a subscriber is?” versus being able to say “Here’s what we mean by a subscriber, and here’s how every system in our estate maps into that.” It matters more than a new BSS because it gives you back control of your business. You build the ontology on top of your existing systems, own the context layer, and make your architecture vendor-agnostic from day one. A new BSS just recreates the same dependency with a different logo on the invoice.
AI doesn’t fail in telco because the models aren’t smart enough. It fails because the models have no coherent view of the business to act on. Every telco system defines “customer,” “product,” and “offer” differently. Without a shared semantic model, your AI is constantly doing archaeology—reverse-engineering what each system means—instead of doing work. When you have a telco ontology, your AI can plug into a consistent, correct view of your business. New use cases reuse the same context instead of redefining it from scratch. That’s what makes AI be able to work in the real world, with your actual architecture.
AI capabilities are compounding on Huang’s‑Law timescales, doubling every year or faster. A three- to five-year migration program is optimizing for a world that won’t exist when you finish. By the time you’ve consolidated onto your chosen vendor’s platform, that vendor’s architecture may already be obsolete relative to what AI can do. The smarter play is to make your architecture indifferent to which vendor comes out ahead. An ontology does that. You can trial a new AI use case in one market without rebuilding context. You can swap out a billing function without re-explaining your entire business to every downstream system. The question stops being “Do we bet the farm on this new vendor?” and starts being “What’s the next small, reversible step we can make on top of the context layer we already own?”
First, call Totogi. We can guide you through the process. We’ll start small, picking a narrow, painful slice of your business—dormant cells, change requests, offer launches—and map it into a telco ontology like the Totogi Ontology. Run AI against that clean model and see what becomes possible. Then do it again. Every process you run through the ontology enriches it: edge cases surface missing business rules, outcomes improve decision logic, and the system compounds in value the more you use it. You don’t need to change your systems to start. You need to build your context. Start there, and the migration, modernization, or whatever transformation you’ve been dreading becomes a series of controlled, reversible steps instead of a bet-the-farm moment.