AI Engineer - Multi-Agent Systems & Production AI

SYNC; Jeddah; On-site; SAR 240,000 - 320,000; Full time

Job summary

Eazli is building an AI-native marketplace in Jeddah and across the region, connecting homeowners with products and services through intelligent agents. We seek an AI Engineer who can design, build, test, and operate production-grade AI agents that interact with real business systems rather than demos.

You’ll work on agent workflows, tool integration, and robust prompt engineering in a fast-paced environment, driving practical AI solutions that scale across multiple domains.

Qualifications

• 4+ years of experience in AI/ML, including Deep Learning and Reinforcement Learning.
• Strong Python engineering experience, preferably Python 3.12+.
• Hands-on experience with agentic frameworks (LangChain, LangGraph, CrewAI).
• Experience building LLM applications with tool/function calling and structured outputs.
• Hands-on experience with OpenAI APIs or similar model providers such as Gemini, Anthropic, or local models.
• Practical experience of agent frameworks or orchestration patterns, such as Agno, LangGraph, CrewAI, AutoGen, or similar.
• Solid understanding of vector databases (Milvus, Pinecone, Qdrant), graph design, and RAG architectures.
• Ability to design robust prompts, agent instructions, guardrails, and evaluation datasets.
• Experience with A2A protocol, AG-UI protocol, MCP, or multi-agent communication systems.
• Experience integrating AI systems with business APIs and databases.
• Strong testing mindset, including unit tests, integration tests, and LLM/agent behavior evaluations.
• Familiarity with Redis, Postgres, object storage, Docker, and cloud-native deployment basics.

Responsibilities

• Build and maintain AI agents using Python, Agno, OpenAI/Gemini/Ollama integrations, and custom tool executors.
• Design agent workflows for customer and vendor use cases such as product search, cart/order management, inventory updates, profile management, service search, market research, and redesigning.
• Implement reliable tool-calling agents that interact with ERP APIs, S3 artifacts, vector databases, and external services.
• Improve Eazli's multi-agent architecture using A2A communication, AG-UI streaming, Redis-based registry/check-ins, and team-based delegation.
• Create and maintain YAML-driven agent configurations, prompts, tool schemas, routing behavior, and workflow judges.
• Build RAG/knowledge features using embeddings, Milvus/vector search, and domain-specific retrieval logic.
• Write agentic tests for persona compliance, tool usage, guardrails, context handling, and expected outputs.
• Work with FastAPI, Pydantic, Redis/Valkey, Postgres, S3/MinIO, Docker, and Helm/Kubernetes-oriented deployment patterns.
• Add observability and debugging support using OpenTelemetry, Phoenix/Arize, Sentry, logs, and structured traces.
• Collaborate with backend teams to integrate AI services (FastAPI experience is a plus)
• Continuously experiment, research, evaluate, and improve AI models and system performance

Skills

Python

AI/ML

LangChain

LLM tooling

OpenAI API

Kubernetes

FastAPI

Postgres

Redis

Testing

Education

Bachelor's degree in CS or related field

Tools

Agno

LangGraph

CrewAI

AutoGen

Milvus

Pinecone

Qdrant

Job description

Eazli is building an AI-native marketplace in Jeddah and across the region, connecting homeowners with products and services through intelligent agents. We seek an AI Engineer who can design, build, test, and operate production-grade AI agents that interact with real business systems rather than demos.

You’ll work on agent workflows, tool integration, and robust prompt engineering in a fast-paced environment, driving practical AI solutions that scale across multiple domains.