Data scientist - Agentic AI
Core
Design, implement, and operationalize agentic workflows for Marvis, Juniper's next-gen AI assistant for network operations, enabling autonomous diagnosis and troubleshooting.
Role type
Senior IC machine-learning engineer (agentic AI)
Builds
Production agentic systems, reasoning pipelines, and tool-calling patterns for network operations
Domain
Generative AI, Network Operations, Cloud Infrastructure
Deliverable
production ML models
Required skills
Agentic orchestration frameworks (LangGraph), MCP server implementation, RAG pipeline design, Kubernetes, Python, LLM API integration, Evaluation frameworks for non-deterministic systems, Statistical and machine learning techniques
Preferred skills
Agent memory systems, Sandboxed code execution environments, Networking domain knowledge, Managed AI/ML services (AWS Bedrock, OpenSearch)
Technologies
LangGraph, LangChain, MCP, Kubernetes, Docker, FastAPI, OpenSearch, AWS Bedrock, Jenkins, GitHub Actions, ArgoCD, LangFuse, Prometheus
Responsibilities
Design and iterate agentic workflows with ReACT loops and dynamic tool selection; Develop and maintain MCP servers and skills; Integrate and optimize LLM capabilities at production scale; Build retrieval and memory services including RAG and vector stores; Design evaluation frameworks for non-deterministic systems; Collaborate with domain experts to formalize networking problems as agentic workflows; Deploy containerized services in Kubernetes and manage CI/CD pipelines; Own observability for agentic workflows including tracing and logging
Seniority
Senior, hands-on IC
