Principal AI Architect (Telecom Autonomous Networks)

S Base Technologies • Basking Ridge, NJ

4d • Onsite

About The Position

As the Lead AI Architect for Telecom, you will design the brain of our next-generation autonomous network. You will build an AI platform where Agentic AI acts as the orchestration layer for network service fulfillment, closed-loop assurance, and predictive maintenance. Your goal is to move the organization from "Human-in-the-loop" to "Human-on-the-loop", ensuring AI agents can safely execute commands across IP, Optical (DWDM/OTN), and Wireless infrastructures.

Requirements

• Artificial Intelligence
• 10 - 12 Years Minimum Experience
• Bachelors Qualification
• Reduction in MTTR (Mean Time to Repair): Speed of autonomous fault correlation and resolution via agentic workflows.
• Provisioning Accuracy: Zero-touch fulfillment success rate without manual intervention or configuration drift.
• Guardrail Efficiency: Percentage of "unsafe" or "hallucinated" network commands intercepted by the safety layer before execution.
• Agentic Efficiency: Token-to-action cost optimization for high-frequency network monitoring tasks

Responsibilities

• Agentic Network Orchestration: Architect multi-agent systems that can take a high-level service request (e.g., "Provision a 100G Wavelength service") and decompose it into specific CLI/API commands across multivendor environments.
• Intent-Based Networking (IBN): Develop reasoning engines that translate natural language business intents into technical configurations using frameworks like LangGraph or CrewAI.
• Troubleshooting Agents: Design "Diagnostic Agents" that can autonomously query telemetry data, correlate alarms, and suggest (or execute) remediation steps for network outages.
• Data Architecture for Telco AI: Real-time Telemetry RAG: Architect a high-throughput Retrieval-Augmented Generation (RAG) pipeline that ingests streaming telemetry, PCAP files, and syslog data into Vector Databases for real-time agent context.
• Network Digital Twin Integration: Connect AI agents to Digital Twins and Graph Databases (Neo4j) to ensure they "understand" physical and logical topology before making routing changes.
• Unified Data Fabric: Build the data architecture necessary to bridge silos between Radio Access Network (RAN), Core, and Edge data.
• Telco-Grade Guardrails & Safety: Deterministic Execution: Implement strict guardrails (e.g., NeMo, Guardrails AI) to ensure agents never execute a command that violates "Golden Configuration" standards.
• SLA-Aware Governance: Architect systems that monitor agent actions against Service Level Agreements (SLAs), ensuring autonomous decisions do not prioritize one customer's traffic at the expense of a higher-priority emergency service.
• Policy Enforcement: Design a "Policy Proxy" layer where every agentic action is validated against a library of telecom regulatory and security compliance rules.

What This Job Offers

• Job Type: Full-time
• Career Level: Principal