AI / LLM Engineer — Agentic Orchestration (LangGraph)
- Hiring from
- Egypt, United Arab Emirates
- Work type
- Hybrid
- Posted
625,993 remote jobs, straight from company career pages
100% free · New jobs every minute
Show job descriptionHide job description
SDG Group is a global enterprise data, analytics and AI consultancy. We design and build data and AI solutions for our clients. You will build the heart of a production Agentic AI platform: an orchestration layer on LangGraph that runs customer journeys as stateful agents, retrieves from governed data products, and calls enterprise APIs as tools. This is production engineering in a client environment, not prototyping.
What you will do
Build and harden the LangGraph orchestration layer: stateful agent flows, tool calling, retrieval over governed data products, and human-in-the-loop handling.
Engineer for production: state persistence, error recovery, idempotent external calls, observability, and loop and cost controls.
Build the evaluation harness (golden sets, task-success, faithfulness, latency and cost) that proves each use case against its value case.
Integrate the layer with customer touchpoints through APIs and connect to an existing agentic integration layer.
What you bring
Strong Python and real LLM application development experience.
Hands-on LangGraph (or LangChain or a comparable agent framework) in or near production; you understand checkpointing and state, and why an in-memory store does not ship.
RAG, prompt and agent design, and working with external APIs as tools.
Experience evaluating LLM and agent systems, not only building them.
Nice to have
LiteLLM, vector search, and structured output or function calling at scale.
Arabic-language NLP; telecom or large-enterprise delivery.
Why SDG Group
Build real agentic systems that reach production at enterprise scale, not demos, with a team that takes evaluation and engineering rigor seriously. We welcome strong candidates regardless of current location, provided they can be based in Cairo and travel as the engagement requires.
