Data Scientist

Insight Global is seeking a Data Scientist IV to support a large telecommunications client on their Infrastructure Intelligence and Analytics (IIA) team. This team develops and operates the data platform and AI agent infrastructure that enables proactive network monitoring and automated investigations across enterprise network operations.

This individual will be responsible for designing, building, deploying, and evaluating AI agents that consume anomaly detection outputs and data from a large-scale data lake to automate investigation, diagnosis, and remediation activities. The role will own the full agent lifecycle from development through production deployment and performance evaluation.

The ideal candidate brings strong software engineering fundamentals, hands-on experience building AI agents programmatically using frameworks such as LangGraph, experience working with LLMs and agentic architectures, and a track record of deploying production-grade AI solutions.

Day-to-Day

• Build and enhance AI agents using LangGraph or similar frameworks.
• Integrate agents with data lakes, anomaly detection outputs, APIs, and real-time data sources.
• Deploy agents into production through CI/CD pipelines.
• Evaluate agent performance using gold-standard datasets and observability tools.
• Improve agent quality through prompt engineering, context engineering, and tool design.
• Collaborate with engineering and operations teams to support autonomous remediation initiatives.

Must-Haves

• 5+ years of Data Science, Machine Learning, Software Engineering, or AI development experience.
• Strong Python and object-oriented programming skills.
• Experience building AI agents programmatically (LangGraph or similar).
• Experience with LLMs, agent orchestration, tool calling, retrieval, and memory patterns.
• Strong SQL skills and experience working with large datasets.
• AWS experience (S3, Athena, etc.).
• Experience deploying production-grade solutions using Git, testing, and CI/CD.
• Experience evaluating AI model/agent performance.

Nice-to-Haves

• LangSmith
• MCP (Model Context Protocol)
• GitLab CI/CD
• Telecommunications or network operations experience
• n8n
• REST APIs
• boto3
• Network telemetry, syslogs, or SNMP exposure