We are looking for an experienced AI Architect to join an international project for one of our key enterprise clients. You will be responsible for defining and leading the architecture of production-grade GenAI and agentic solutions (LLM + RAG + deterministic Python pipelines) integrated with complex enterprise platforms.
This is a hands-on engineering and orchestration role focused on integrating LLM capabilities into production systems, rather than a model-training or ML research role.
Project: International enterprise project. Languages: Polish (Native / Fluent) & English (C1)
Key Responsibilities:
Define the target architecture for agentic AI services, including orchestration patterns, tool/function schemas, multi-agent/sub-agent designs, and human-in-the-loop workflows.
Design graph-based stateful orchestration (LangGraph) with durable checkpointing and A2A-pluggable agents, as well as build and consume MCP tool servers.
Architect and continuously optimize RAG pipelines (ingestion, chunking, embeddings, hybrid/vector retrieval, grounding, and evaluation).
Design secure, robust integrations with enterprise applications via REST APIs and event-driven patterns (OAuth2 / SSO / Azure AD).
Establish standards for quality, reliability, and responsible AI: evaluation harnesses, guardrails, observability, cost/latency control, fallbacks, and quotas.
Drive architectural alignment across teams, mentor engineers, and build strong foundations for