Data Scientist - Agentic AI

Description

The role is hybrid, with an expectation of about 2 days per week in a company office.

The company works from edge to cloud, helping customers connect, protect, analyze and act on their data and applications.

The Data Scientist - Agentic AI builds and runs the core agentic workflows behind Marvis, Juniper's AI assistant for network operations.

The work sits between data science, generative AI and production engineering. The goal is an AI system that diagnoses, troubleshoots and resolves complex networking problems on its own, and that operates reliably and can be observed in a cloud-native setting.

The ideal candidate has hands-on experience with LLM agent frameworks, preferably LangGraph, and the engineering discipline to ship reliable, observable AI systems.

This is a management-level role. You are a subject matter expert in agentic AI and applied GenAI, you influence technical components of products and services on an ongoing basis, and you may lead projects and mentor colleagues.

You will work with product and engineering teams and use your own judgment to decide how to reach objectives.

Responsibilities

• Build, refine and test agentic workflows in LangGraph, including ReAct loops, dynamic tool selection and binding, multi-step reasoning and self-correction.
• Create and maintain MCP (Model Context Protocol) servers and skills: define tool schemas, build domain tools, write SKILL.md playbooks, and handle versioning, deployment and monitoring.
• Make LLM calls production-ready, covering structured outputs, streaming, tool calling, prompt design and error handling across agent execution paths.
• Build and tune retrieval and memory services, including RAG pipelines, vector search, hybrid retrieval, and semantic, episodic and procedural agent memory.
• Design and run evaluation frameworks for non-deterministic agents: set metrics, build test harnesses, run A/B tests on skills and tool setups, and keep improving quality.
• Work with network engineers and product managers to turn troubleshooting playbooks into agent skills, tools and data pipelines.
• Write data analysis and transformation code that runs in sandboxed Code Mode environments, including multi-tool orchestration, aggregation and visualization.
• Deploy and operate containerized services on Kubernetes, and contribute to CI/CD, container images, health probes and resource tuning.
• Add tracing, logging, cost tracking and performance monitoring to agentic workflows so that these systems stay reliable in production.

Requirements

• A bachelor's, master's or PhD in computer science, data science, mathematics, statistics, engineering, information systems or a closely related quantitative field.
• Typically 4 to 7 years building production ML or AI systems, including at least one year working hands-on with generative AI and LLM applications.
• Production experience with agent orchestration frameworks such as LangGraph (strongly preferred), LangChain, Claude Agent SDK or similar, beyond prototypes.
• Working knowledge of agent design patterns: ReAct loops, tool and function calling, dynamic tool binding, skill-based execution, multi-step planning and self-correction.
• Hands-on experience with MCP or similar tool-serving protocols, including tool schema design, server implementation and registry management.
• Experience integrating LLM APIs at scale, covering prompt engineering, structured outputs, streaming, error handling and cost control.
• Experience designing RAG pipelines: chunking, re-ranking, hybrid search, vector stores such as OpenSearch, and relevance tuning.
• Experience building evaluation and test frameworks for non-deterministic AI systems, including offline evals, A/B testing and LLM-as-judge.
• Strong grounding in statistics and machine learning, such as anomaly detection, time-series analysis, clustering or causal inference.
• Practical judgment on when to use retrieval versus fine-tuning, and prompt engineering versus structured generation, plus the ability to debug model behavior in production.
• Proficient Python developer with experience in production codebases, not only notebooks.
• Kubernetes experience deploying, scaling and managing workloads, including Deployments, Services, ConfigMaps, Secrets and health probes.
• Experience with CI/CD pipelines for build, test and deploy, using tools such as Jenkins, GitHub Actions or ArgoCD.
• Container image management with Docker, including building, tagging and versioning, and familiarity with a registry such as ECR or GCR.
• Backend development with FastAPI or similar, and REST or GraphQL API design.
• Experience with observability tools for AI systems, such as LangFuse or Prometheus, covering tracing, monitoring and logging.
• Preferred: experience with agent memory systems such as LangMem or custom memory designs.
• Preferred: familiarity with sandboxed code execution environments such as E2B or Firecracker.
• Preferred but not required: networking knowledge, including wireless and wired diagnostics and network troubleshooting.
• Preferred: experience with AWS Bedrock, OpenSearch Serverless or similar managed AI and ML services.
• Preferred: strong written and verbal communication, including explaining technical designs to senior leadership.

Conditions

Hybrid role with an average of 2 days per week in a company office.

Job family: Engineering. Job level: TCP03.

The company is an equal opportunity employer and does not discriminate on the basis of race, gender or any other protected category.

Candidates with disabilities who need accommodation during hiring can request it through the company's secure form.

Benefits

• Health and wellbeing benefits covering physical, financial and emotional wellbeing.
• Career development programs, including paths to become a recognized expert in your field or to move into another division.
• Flexibility to manage work and personal needs.