Mosaic One Agents

AI agent · Workflow / back-office agent · Operations & supply chain by Adtran

Scale telecom expertise with role-specific AI agent teams

A suite of role-specific AI agents for telecom network management, customer care and business operations. It enables organizations to scale expertise by automating investigation and coordination across multi-vendor networks while maintaining human control.

Agent specs

The full spec card (context, pricing per token, modalities and more) on AntSpace. Sign up free to see the full spec card.

Launched
2026-10-08

Category: AI Agents

Data sources
Adtran and third-party hardware telemetry
Models
Mosaic One Fabric
Foundation platform
Mosaic One Fabric
Target industries
Telecommunications, Service Providers
Hardware compatibility
Adtran and third-party hardware
  • Role-specific AI agent teams for network, care and operations
  • Normalizes telemetry across Adtran and third-party hardware
  • Built-in governance and runtime guardrails for safe operation
  • Traceable actions with user-defined levels of autonomy

Overview

Mosaic One Agents is a suite of role-specific AI agents designed for telecommunications organizations to scale expertise across network management, customer care and business operations. Built on the Mosaic One Fabric foundation, these agent teams access data from both Adtran and third-party networks to investigate issues in parallel. The system handles time-consuming coordination and process tasks, allowing human experts to direct multiple workflows simultaneously rather than managing every task manually. It provides an operating model where people set the intent and AI agents move work forward under human direction and governance. By normalizing telemetry across multi-vendor hardware, the agents provide a unified view across network domains to drive issues through to resolution.

Operational Intelligence

  • Role-specific teams: The suite includes role-specific AI agent teams tailored for network management, customer care and business operations.
  • Coordinated investigation: Agents draw on shared intelligence from the Mosaic One Fabric platform to investigate issues and coordinate work across teams.
  • Multi-vendor normalization: The system normalizes telemetry across Adtran and third-party hardware to provide a unified view across network domains.
  • Deterministic intelligence: Agents use root-cause analysis, cross-domain correlation and service-impact assessment to evaluate options and advance workflows.

Governance and Control

  • Defined autonomy: Users define the level of autonomy for each use case, ranging from surfacing recommendations to staging actions for approval.
  • Safety guardrails: The platform includes built-in governance and runtime guardrails to keep the network and agents safe.
  • Traceable actions: Every step taken by the AI agents remains visible and traceable for human review and oversight.
  • Scalable autonomy: Organizations can increase agent autonomy over time as confidence in their performance grows.

Use cases

  • Network engineers use agent teams to investigate outages in parallel across Adtran and third-party hardware domains.
  • Customer care teams leverage shared intelligence to resolve subscriber issues faster by automating coordination tasks.
  • Operations managers set intent-based guardrails to allow AI agents to stage network actions for human approval.
  • Service providers scale their expert workflows by directing multiple AI-driven processes simultaneously during major outages.

Specifications

Data sources
Adtran and third-party hardware telemetry
Models
Mosaic One Fabric
Foundation platform
Mosaic One Fabric
Target industries
Telecommunications, Service Providers
Hardware compatibility
Adtran and third-party hardware
Intelligence types
Root-cause analysis, cross-domain correlation, service-impact assessment
Control modes
Recommendations, staged actions for approval, predefined boundaries

How it works

  1. Data Integration: The agents access normalized telemetry from Adtran and third-party networks via the Mosaic One Fabric foundation.
  2. Parallel Investigation: Role-specific agent teams investigate problems in parallel, using deterministic intelligence like root-cause analysis.
  3. Human-Directed Workflow: Human experts set the intent and direct multiple workflows, while agents handle coordination and process tasks.
  4. Governed Execution: Agents stage actions for approval or operate within predefined boundaries set by the organization's governance rules.

FAQ

What is the relationship between Mosaic One Agents and Mosaic One Fabric?

Mosaic One Agents is built on Mosaic One Fabric, which provides the open foundation for building and governing AI agents around an organization's data and networks.

Does it work with non-Adtran hardware?

The agents use normalized telemetry from both Adtran and third-party hardware, providing a unified view across different network domains.

How do organizations maintain control over AI actions?

Users retain control through built-in governance and runtime guardrails, allowing them to define the level of autonomy for each use case, such as requiring approval for staged actions.

What specific roles do these AI agents support?

The suite provides role-specific teams for network management, customer care and business operations.

Added by AntSpace from public sources (adtran.com).

Related

More AI agents by Adtran