AWC Miami: The Two Camps of Telecom and Why AI Is About to Split Them Apart

I just got back from AWC Miami with the BTS team. I go to these conferences for the voice interconnect business — meetings, handshakes, the usual wholesale telecom circuit. But this year, something was different.

Every conversation I had eventually landed on the same topic: AI.

Not in the breathless, Silicon Valley way. In the practical, "what does this actually mean for my business" way. And what I heard split cleanly into two camps that I think define where this industry is headed.

Camp One: The Experimenters

These are the carriers who are already playing with it. Some are using Claude Code to automate internal processes. One carrier I sat down with is building a full-stack AI voice solution — not a chatbot wrapper, not a demo, actual production infrastructure for opportunities that he is seeing in the automated voice customer experience space.

They're not all technical. That's the part that surprised me most.

I sat in meetings with carriers who have never written a line of code in their lives, and they were genuinely experimenting. One guy was building Polymarket arbitrage bots. Another had an AI assistant drafting customer communications. Most of it was messy. Some of it was frankly ridiculous. But they were in the arena, getting their hands dirty, figuring out what works and what doesn't.

The quality of what they're building isn't the point. The fact that they're building at all is the point. Because experimentation compounds. The person who spent six months fumbling with AI tools is going to be dramatically more capable than the person who starts from zero when it finally becomes unavoidable.

And it will become unavoidable.

Camp Two: The Legacy Play

Then there's the other camp. The carriers who aren't thinking about AI at all.

These tend to be larger, more established operations. They have deep customer relationships built over decades. They have reliable revenue streams. They have processes that work well enough. And they're betting — whether they realize it or not — that those relationships and that institutional momentum will protect them from technological disruption.

I've seen this movie before. I've been building companies since 1998. I watched e-commerce disrupt print catalogs. I watched streaming disrupt physical media distribution. I built one of the first live commerce platforms in 2013, years before "live shopping" became a category. Every time, the incumbents said the same thing: "our relationships are too strong, our customers are too loyal, this doesn't apply to our industry."

It always applies.

The carriers relying on legacy relationships aren't wrong that relationships matter. They're wrong that relationships alone are a moat. When a competitor can do in minutes what takes your team hours — reconciling CDRs, analyzing rate decks, optimizing routing — the relationship advantage erodes fast. Customers are loyal until someone else is meaningfully better.

The Real Gap

The gap in telecom right now isn't between big carriers and small carriers. It's not between technical teams and non-technical teams. It's between the people who are paying attention and the people who aren't.

The non-technical carrier owner fumbling with a Polymarket bot is closer to the future than the established carrier with a 50-person operations team who hasn't looked at any of this. Because the fumbler is building intuition. He's learning what AI can and can't do, where it breaks, where it's magic. That intuition will be worth more than any specific tool he builds along the way.

I know this because it's exactly what happened to me.

What I'm Actually Doing About It

I work in wholesale VoIP operations by day. Nights and weekends, I build tools that solve the operational problems I see every day at work.

CDRCheck handles billing reconciliation — upload two CDR files, match millions of records, find every discrepancy. VoIP Accelerator does the same for rate decks — compare 250K+ records and see exactly where rates diverged. Neither tool uses AI internally. Both were built in weeks using AI coding assistants — work that would have taken months solo.

I also run an autonomous AI agent system on a Mac Mini that handles research, monitoring, and content drafting 24/7. Three specialized agents, scheduled workflows, persistent memory across sessions. The infrastructure isn't the product — it's the force multiplier that lets me ship at a pace that would otherwise require a team.

This isn't theoretical. These tools are live. Carriers use them. And I built all of it while working a full-time job, because the tools to do that are better than they've ever been.

Anthropic Thinks Telecom Matters

One more thing. I just signed up for the Claude Partner Network. When you register, there are about seven industry categories you can select.

One of them is Telecom.

That's not an accident. Anthropic — the company behind Claude, one of the most capable AI systems on the planet — is explicitly building a partner ecosystem for telecom. I'm also working toward the Claude Certified Architect certification, because if I'm going to build in this space, I want to know the platform at an architectural level, not just as a user.

They see what's coming. The question is whether the industry sees it too.

What This Means If You're in the Space

If you're already experimenting: keep going. The fumbling phase is temporary. The intuition you're building is permanent. Find the specific operational pain points in your business and point AI at them. Not the sexy stuff — the boring stuff. The reconciliation, the rate analysis, the routing optimization. That's where the real leverage is.

If you're not experimenting yet: start now. Not next quarter. Not after the next board meeting. Now. Pick one process that eats hours every week and see what happens when you throw Claude Code at it. You'll be surprised how much of your "this is just how we do it" workflow is actually unnecessary friction.

If you're relying on legacy relationships to protect your business: those relationships are real, and they matter. But they're not a strategy. They're a buffer. And buffers run out.

The carriers who figure this out in the next 18 months will have a structural advantage that compounds over time. The ones who don't will spend the next five years wondering what happened.

I know which camp I'm in.