Agentic AI Engineer - Product Support

Location: Tel Aviv District, Israel

Company

Building a cloud-delivered platform that converges enterprise networking and security into a centralized SASE service.

What you will do

• Identify automation and optimization opportunities with Product Support and other business teams.
• Design and implement agent-based AI workflows, copilots, and LLM-powered tools using orchestration frameworks such as LangGraph and LangChain.
• Develop RAG pipelines, integrate multiple data sources, and validate data quality for AI solutions.
• Take AI solutions from ideation and prototyping through production deployment.
• Implement guardrails, approvals, logging, human-in-the-loop controls, and monitoring for reliability, security, privacy, compliance, hallucinations, and drift.
• Document AI flows and operational runbooks, translate business requirements into solutions, and educate internal teams on AI capabilities and limitations.

Requirements

• At least 3 years of hands-on software development experience delivering and maintaining production-quality code.
• At least 2 years of hands-on GenAI development experience with a strong AI/ML focus.
• Strong Python skills and a production mindset.
• Deep understanding of LLMs, SLMs, prompt engineering, context engineering, and agent-based systems.
• Experience with LangGraph, crewAI, Strands, or similar orchestration frameworks, including bringing enterprise-grade agents into production.
• Understanding of ML workflows and MLOps, with strong analytical skills and the ability to work directly with non-technical stakeholders.

Nice to have

• Experience with AWS Agent Core.
• Experience with agentic AI applications such as n8n, UiPath, or Make.
• Experience or familiarity with the Product Support domain.

Culture & Benefits

• Hands-on builder role focused on practical, high-impact AI solutions rather than research.
• Close collaboration with business owners and stakeholders across the organization.
• Opportunity to build custom solutions where off-the-shelf tools are insufficient.