- Jobs
- NTT DATA Romania
- AI Engineer
AI Engineer
AI Tools
Agentic Frameworks
Tech Stack
Agent Workflow
Lead implementation of multi-agent systems, design complex agentic systems with multiple interacting agents, implement orchestration logic with state machines/retries/fallbacks/human escalation, implement RAG pipelines and tool calling with Google ADK and LangGraph
About the Role
NTT DATA Romania - AI Engineer Multiple locations: Cluj, Timisoara, Iasi, Brasov, Sibiu, Bucharest, Romania (Remote options available)
Lead the implementation, deployment and test of multi-agent systems. Design and build complex agentic systems with multiple interacting agents. Implement orchestration logic including state machines, retries, fallbacks, and human escalation. Implement RAG pipelines, tool calling, and sophisticated system prompts for optimal reliability. Apply ML concepts to evaluate and improve agent performance.
Key Responsibilities:
- Lead development using Google ADK and/or LangGraph
- Integrate supporting libraries (LangChain, LlamaIndex, vector databases, message queues)
- Build evaluation suites and run regression testing
- Deploy AI services with CI/CD pipelines and observability
- Debug complex failures end-to-end
- Collaborate with product managers on requirements
- LLM-powered agents and multi-agent workflows
Requirements:
- Bachelor's in Computer Science or related field
- 3+ years as Software/ML/AI Engineer
- 1-2 years directly working with LLMs in production
- Strong Python proficiency (async, type hints, testing)
- Git, unit/integration testing, CI/CD expertise
- REST/gRPC API design and integration
- Deep LLM knowledge: tokenization, context windows, temperature, prompting patterns
- Prompt engineering patterns (ReAct, chain-of-thought, tool-calling/tool-use)
- Fine-tuning, adapters, RAG experience
- Experience building LLM-powered applications end-to-end
- Hands-on with Google ADK or LangGraph
- Vector database experience (Pinecone, Weaviate, pgvector, Chroma)
- Cloud deployment (GCP, AWS, Azure)
Nice-to-have:
- Vertex AI/Gemini experience
- LangChain, LlamaIndex, semantic search
- Monitoring stacks (OpenTelemetry, Prometheus, Datadog)