Agentic Telecom Glossary | Decoding AI and Network Planning Terms

Agentic Telecom Glossary

Decoding the terms used in AI and telecommunications.

Agentic AI

A category of artificial intelligence where autonomous agents perceive their environment, make decisions, and take actions to achieve specific goals without requiring step-by-step human instructions. In telecommunications, agentic AI handles complex planning, optimization, and operational tasks independently.

AI Agent

An AI agent is a software program designed to autonomously perform tasks by perceiving its environment, analyzing data, and taking actions to achieve specific goals. AI agents operate independently within defined parameters, using machine learning, optimization algorithms, and other advanced technologies to execute tasks without requiring continuous human intervention.

As-Built Documentation

Records showing the actual installed configuration of a network, including any deviations from the original design. Accurate as-built documentation is critical for network maintenance, troubleshooting, and future expansion planning.

Autonomous Agent

A self-governing software program capable of making decisions and performing tasks without direct human oversight. In network planning, autonomous agents can design routes, optimize material usage, and dispatch crews independently.

Backhaul

The portion of a network that connects the access network (serving end users) to the core network. In FTTH deployments, backhaul typically refers to the fiber connections between distribution hubs and central offices.

Bill of Materials (BOM)

A comprehensive list of all components, materials, and quantities required to build a network design. AI agents can automatically generate accurate BOMs from network designs, eliminating manual counting errors.

Cable Route Optimization

The process of determining the most efficient path for fiber cables considering factors like distance, terrain, existing infrastructure, permits, and installation costs. AI agents use mathematical optimization to evaluate thousands of possible routes.

Central Office (CO)

A facility housing telecommunications equipment that serves as a hub for network connections. In FTTH networks, the central office typically contains the Optical Line Terminal (OLT) equipment.

Connector

In the context of AI platforms, connectors are API integrations with a customer's existing tools, allowing AI agents to access, query, and enhance their understanding of network data, project status, and operational context.

Cortex

Nexma's AI-powered operations software that orchestrates crew dispatch, project tracking, and predictive deployment analytics. Cortex agents handle scheduling, resource allocation, and operational optimization.

Demand Clustering

The process of grouping customer locations to determine optimal placement of network distribution points like splitters and cabinets. AI agents analyze geographic and infrastructure data to identify mathematically optimal cluster configurations.

Drop Cable

The final segment of fiber connecting a distribution point to an individual customer premise. Drop cable routing and installation represent a significant portion of FTTH deployment costs.

Feeder Cable

High-capacity fiber cables that carry traffic from the central office to distribution points in the network. Feeder cable routing is a critical factor in overall network cost optimization.

Fiber Distribution Hub (FDH)

An enclosure in the outside plant where feeder cables connect to distribution cables. FDH placement optimization is a key function of network planning AI agents.

Fiber to the Home (FTTH)

A network architecture delivering fiber optic connectivity directly to residential or business premises, providing high-bandwidth internet, voice, and video services.

Field Survey

The process of collecting physical site data needed for network design, including pole locations, conduit routes, building access points, and terrain conditions. AI agents can process survey data to automatically generate network designs.

Forge

Nexma's AI-powered field execution platform enabling voice-guided installation and real-time quality verification. Forge agents assist technicians with installation procedures and validate work against design specifications.

Foundation Model

A large-scale machine learning model trained on massive datasets that can be fine-tuned for specific tasks. Nexma's agents leverage foundation models enhanced with telecommunications domain expertise.

Generative AI (GenAI)

A type of AI that creates new content based on patterns in training data. In network planning, generative AI can produce complete network designs from input parameters and constraints.

Geographic Information System (GIS)

Software for capturing, storing, analyzing, and visualizing geographic data. GIS platforms are foundational to network planning, providing the spatial context AI agents need to design optimal routes.

Geospatial Intelligence

The analysis and visualization of location-based data to derive actionable insights. Nexma's platform fuses GIS, 3D visualization, and infrastructure data to enable precise network design.

Handhole

An underground access point for fiber cables, smaller than a manhole, used for cable pulling, splicing, and maintenance access.

Last Mile

The final leg of a telecommunications network connecting service providers to end customers. Last mile infrastructure, particularly in FTTH, represents the largest deployment cost and complexity.

Mixed Integer Programming (MIP)

A mathematical optimization technique used to solve complex planning problems with both continuous and discrete variables. Nexma's agents use MIP to find mathematically optimal network designs.

Network Design Automation

The use of software and AI to automatically generate network plans from input data, replacing manual CAD work with algorithmic optimization.

Nexma Agents

AI agents built by Nexma.ai specifically for telecommunications network planning and deployment. These agents are experts in their tasks and trained to deliver optimal network designs and operational outcomes.

Nexus

Nexma's AI-powered geospatial planning software that transforms field surveys into optimized network designs. Nexus agents handle route optimization, splice point placement, and material calculations.

Non-Human Work

Repetitive tasks that do not require human creativity or strategic thought, ideal for automation by AI agents. In telecommunications, this includes manual route tracing, material counting, and schedule coordination.

Optical Line Terminal (OLT)

Equipment located in the central office that serves as the service provider endpoint of a passive optical network, managing downstream and upstream traffic to customer premises.

Optical Network Terminal (ONT)

Customer premises equipment that terminates the fiber connection and converts optical signals to electrical signals for customer devices.

Optimization Algorithm

A mathematical procedure for finding the best solution from a set of possible options given specific constraints and objectives. Network planning AI agents use optimization algorithms to minimize costs while meeting service requirements.

Outside Plant (OSP)

The physical infrastructure of a telecommunications network located outside of buildings, including cables, poles, conduits, handholes, and cabinets.

Passive Optical Network (PON)

A fiber network architecture using unpowered optical splitters to deliver service from a single fiber to multiple customer premises, reducing infrastructure costs.

Permit Management

The process of obtaining necessary approvals from municipalities, utilities, and property owners for network construction. AI agents can cross-reference designs against permit requirements to identify compliance issues.

Right-of-Way (ROW)

Legal permission to install network infrastructure along a specific route, typically on public property or utility easements. ROW availability is a critical constraint in network route optimization.

Route Optimization Agent

An AI agent that evaluates possible cable paths and applies mathematical optimization to find designs that minimize material costs and installation complexity.

Service Area

A defined geographic region targeted for network deployment. AI agents can analyze service areas to determine optimal phasing and infrastructure requirements.

Slack Loop

Extra cable length coiled at specific points in the network to allow for future maintenance, repairs, or network modifications without requiring new cable pulls.

Span

The section of aerial cable between two consecutive poles. Span length affects cable sag, tension, and wind loading calculations.

Splice Point

A location where fiber cables are joined together, either through fusion splicing or mechanical connectors. Optimal splice point placement reduces material waste and installation time.

Splitter

A passive optical device that divides a single fiber signal into multiple outputs, enabling one feeder fiber to serve multiple customers in PON architectures.

Strand Map

A detailed diagram showing pole locations, span lengths, existing attachments, and proposed cable routes for aerial network construction.

Work Order

A document authorizing and detailing specific work to be performed, including location, materials, procedures, and scheduling. AI agents can automatically generate work orders from approved designs.

‍