AI Architect: LLM Systems & Multi-Agent Design

Nitor Group Oy; Helsinki; On-site; EUR 90,000 - 120,000; Full time

Job summary

Nitor in Helsinki is seeking a senior AI architect to design and build scalable AI systems that bridge business strategy and engineering. You will lead architecture initiatives and mentor teams to implement robust LLM-based solutions.

You will drive development of memory-aware, RAG-enabled workflows and multi-agent configurations, ensuring production-grade reliability across tools and platforms.

Qualifications

• Experience designing scalable AI systems and architectures.
• Deep understanding of LLM-based system design including memory and context graphs.
• Knowledge of agent frameworks and integration with AI platforms.
• Familiarity with MLOps practices for testing, monitoring and observability.

Responsibilities

• Designing and building scalable AI systems at Nitor, bridging business strategy and AI engineering, mentoring, and developing AI architecture practices.
• Deep understanding of LLM‑based system design, including RAG architectures, multi‑agent systems, tool use, and balancing deterministic and probabilistic workflows.
• Experience with agent frameworks such as LangGraph, LangChain, OpenAI Agents SDK, Google ADK, or Claude Agent SDK.
• Knowledge of AI platforms such as Microsoft Foundry, AWS Bedrock, or Gemini Enterprise Agent Platform.
• Experience with agent memory architectures and context graphs, familiarity with vector search and graph database integration.
• Understanding of LLMOps and EvalOps practices, including testing, monitoring, and observability of AI systems in production.

Skills

LLM design

RAG architecture

Multi-agent systems

Tool integration

Tools

LangGraph

LangChain

OpenAI Agents SDK

Google ADK

Claude Agent SDK

Microsoft Foundry

AWS Bedrock

Gemini Platform

Job description

Nitor in Helsinki is seeking a senior AI architect to design and build scalable AI systems that bridge business strategy and engineering. You will lead architecture initiatives and mentor teams to implement robust LLM-based solutions.

You will drive development of memory-aware, RAG-enabled workflows and multi-agent configurations, ensuring production-grade reliability across tools and platforms.