Job Summary
Agentic AI Developer - Telecom Domain
Key Responsibilities
Design, build, test, and productionize telecom-domain agents for Nokia’s Agentic Control Plane across 5G Core, Packet Core, IMS, and Autonomous Networks.
Skill Requirements
Build production-grade AI Agents using LangGraph, LangChain, Semantic Kernel, CrewAI, AutoGen, and LlamaIndex.
Design Agentic RAG solutions over telecom documentation, alarms, KPIs, topology, inventory, and operational runbooks.
Implement ingestion, chunking, embeddings, vector retrieval, reranking, grounding, and citation generation.
Develop AI microservices using Python, FastAPI, REST APIs, AsyncIO, and containerized architectures.
Create MCP Servers and MCP Clients exposing approved telecom tools, repositories, databases, and APIs.
Integrate LLMs through Vertex AI and enterprise model gateways with prompt caching and model routing.
Develop multi-step agent workflows with tool calling, contextual memory, state management, and human approvals.
Apply 5G Core, IMS, EPC, OSS/BSS, assurance, and operational domain expertise to agent design.
Design structured tool contracts and agent interfaces for reusable telecom capabilities.
Build evaluation pipelines covering grounding, telecom correctness, hallucination detection, latency, and safety.
Implement observability for prompts, retrieval quality, tool invocations, token usage, and business outcomes.
Integrate agents with PostgreSQL, vector databases, Kafka/PubSub, enterprise APIs, and telemetry platforms.
Implement Responsible AI controls, access controls, auditability, approval workflows, and security guardrails.
Deploy solutions on GKE/Cloud Run using Docker, GitLab CI/CD, automated testing, and release pipelines.
Own production rollout, incident analysis, operational readiness, regression testing, and service evolution.