Saperly Launches as the First Phone Carrier Built for AI Agents
Krasa AI
2026-05-05
4 minute read
Saperly Launches as the First Phone Carrier Built for AI Agents
Here's a question nobody was asking two years ago: what phone carrier does your AI agent use? Saperly, a new startup that launched today, thinks this is the next billion-dollar infrastructure problem — and they might be right.
Saperly is positioning itself as the first phone carrier purpose-built for AI agents. It provides real phone numbers that any MCP-compatible (Model Context Protocol) agent can use for calling and texting, complete with a stable identity across products, workflows, and channels.
The Problem Saperly Solves
Right now, if you want your AI agent to make phone calls or send texts, you're cobbling together solutions from carriers designed for humans or developer APIs that weren't built for autonomous systems. The result is fragile, compliance-heavy, and doesn't scale well.
Consider what an AI agent actually needs from a phone carrier. It needs a persistent identity (so customers recognize the number). It needs the ability to handle multiple concurrent conversations. It needs compliance and audit logging built in, not bolted on. And it needs to work seamlessly with the agent frameworks that developers are already using.
Traditional carriers don't offer any of this. Twilio and similar APIs get you partway there, but they're tools for developers building voice apps — not infrastructure for autonomous agents that need their own communication layer.
How It Works
Saperly provides a unified phone number for each agent that works for both calling and messaging. The key differentiator is MCP compatibility — any agent built with the Model Context Protocol (the increasingly standard way AI agents connect to external tools) can plug into Saperly natively.
The platform handles three core functions. First, identity management: agents get persistent phone numbers that follow them across products and workflows. Second, communication: both voice calls and text messages through standard phone infrastructure. Third, compliance: disclosure, consent, and audit logging can be enabled per line, so compliance requirements don't have to live in application code.
That last point matters more than it sounds. As AI agents increasingly interact with real people over the phone, regulations around disclosure ("you're speaking with an AI") and consent are tightening globally. Building that into the carrier layer means developers don't have to solve it themselves for every agent they deploy.
Why This Matters Now
The timing makes sense when you look at the broader AI agent landscape. MCP adoption has exploded in 2026 — the Linux Foundation reported over 10,000 active public MCP servers and tens of millions of monthly SDK downloads. Agents are no longer experiments; they're production systems handling real workflows.
Many of those workflows involve communicating with humans. Sales agents qualify leads over the phone. Support agents handle customer calls. Scheduling agents confirm appointments via text. Each of these needs reliable, compliant phone infrastructure.
Until now, the agent stack had a clear gap between "AI reasoning" and "real-world communication." Saperly fills that gap by treating phone infrastructure as a first-class requirement for autonomous systems, not an afterthought.
The Competitive Landscape
Saperly enters a space that's starting to heat up. OpenAI recently revealed plans for an AI-native smartphone where agents replace traditional apps. Google's Agent2Agent protocol enables cross-platform agent communication. The infrastructure layer for AI agents is being built out rapidly.
But Saperly is approaching from a unique angle — the telecom layer. Rather than competing with agent frameworks or LLM providers, they're providing the communication plumbing that every agent needs but nobody has purpose-built before.
What to Watch
The key question for Saperly is adoption velocity. If major agent platforms integrate Saperly as their default communication layer, the company could become essential infrastructure. If developers stick with existing APIs and build compliance in-house, it'll be a harder sell.
The bottom line: as AI agents move from chatbots to autonomous systems that interact with the real world, they need real-world infrastructure — including their own phone numbers. Saperly is betting that "phone carrier for AI agents" is a category that's about to explode. Given where the industry is heading, that bet looks increasingly smart.
Don't fall behind
Expert AI Implementation →Related Articles
NVIDIA Cosmos 3: First Open Physical AI Omnimodel Cuts Training Cycles to Days
NVIDIA's Cosmos 3 launches at Computex 2026 — a fully open foundation model that unifies vision, world generation, and action for robots and autonomous systems.
min read
Anthropic Adds Services Track and Partner Hub to Claude Network
Anthropic launches a 3-tier Services Track and a public Partner Hub. 40,000 firms have applied; 10,000 consultants are certified.
min read
Apoha Exits Stealth With $36M to Build 'Liquid Brain' AI for Materials
UK startup Apoha emerges with $36M Series A and a wild new data type: how materials vibrate in liquid. The pitch is AI for materials discovery.
min read