AI Agent Architect

Lucidya is the leading customer experience intelligence platform in the MENA region, helping organizations deeply understand and serve their customers using cutting-edge AI. As we expand our AI capabilities, we are building agentic AI services that can reason, plan, and act autonomously to deliver real business value.

We are looking for an AI Agent Architect to design and lead the architecture of our next-generation AI agent systems.

As an AI Agent Architect, you will own the end-to-end design of agent systems at Lucidya. Your role is to define how agents reason, interact with tools, collaborate as multi-agent systems, manage memory, and operate safely at scale. You will work closely with ML engineers, backend engineers, product managers, and UX designers to turn advances in LLMs and agent frameworks into robust, production-ready systems fully integrated with Lucidya’s platform.

We are seeking someone with a deep understanding of AI technologies who can architect systems that enable intelligent, goal-driven behaviors in our products. Collaboration is key: you will work across teams to align AI solutions with business needs, enhance user experiences, and bring innovative ideas from concept to production.

Key Tasks & Responsibilities

• Design and implement the architecture of Lucidya’s AI agents, including reasoning loops, planning, reflection, and multi-agent collaboration.
• Research and apply state-of-the-art methods in machine learning, natural language processing, and reinforcement learning.
• Guide the development of prototypes and proof-of-concept models, experimenting with emerging technologies.
• Collaborate with software engineers and data scientists to integrate AI capabilities seamlessly into Lucidya products.
• Architect scalable and secure agent–tool interactions (e.g., CRM, analytics APIs, external systems).
• Ensure AI solutions align with product goals and deliver measurable business value.
• Design short-term and long-term memory architectures for agents.
• Build retrieval-augmented generation (RAG) pipelines using vector databases (Pinecone, Weaviate, FAISS).
• Conduct evaluations of agent performance and optimize algorithms based on analytics and user feedback.
• Define observability standards for monitoring, debugging, and improving agent behaviors.
• Produce clear design documentation, specifications, and best practices for AI agent development.
• Establish AI governance and safety guardrails, ensuring agents act responsibly and in compliance with regulations.
• Stay current on the latest advancements in LLM, multi-agent, and AI systems research.
• Mentor engineers on agent orchestration and AI architecture best practices.
• Act as the subject matter expert and key driver of agentic AI within Lucidya.