About the Role

We are looking for a highly skilled and hands‑on AI Architect to design and lead the implementation of multi‑agent (agentic) systems in enterprise environments.

This is not a prompt engineering or chatbot role. We are focused on building production‑grade AI systems, where multiple agents collaborate, reason, and execute complex workflows integrated with real business processes.

Key Responsibilities
- Design multi‑agent architectures (task decomposition, orchestration, coordination patterns).
- Define how LLM‑powered agents interact with enterprise data platforms, APIs and tools, operational workflows.
- Lead agentic systems from PoC to production with model, cost, security, privacy, and responsible AI guardrails.
- Establish observability, tracing, feedback loops, and controls for agent behavior.
- Define memory strategies (short‑term, long‑term, contextual grounding).
- Collaborate with data, platform, and engineering teams to integrate AI into core systems.
- Guide and train teams on best practices for scalable and reliable AI systems.

Qualifications





What we’re looking for
- Strong background in software architecture and distributed systems.
- Hands‑on experience building complex LLM‑based applications.
- Experience designing complex workflows or orchestration systems.
- Solid understanding of RAG architectures, retrieval optimization, and retrieval quality.
- Strong understanding of LLM fundamentals and transformer architecture.
- Solid grasp of transformer attention, embeddings, tokenization.
- Understanding of LLM training paradigms (pretraining vs fine‑tuning vs instruction tuning).
- Knowledge of context windows, limitations, hallucinations and mitigation strategies.
- Familiarity with NLP concepts such as semantic similarity, text embeddings, vector representations.
- Information retrieval principles and ability to reason about model behavior and limitations.
- Experience with multi‑agent frameworks (LangChain, Semantic Kernel, Agent Fram

📌 AI Architect – Agentic Systems (Medellín)
🏢 Endava
📍 Medellín