Senior Data Scientist

We are a provider of digital transformation and technology consulting services with a portfolio of solutions for both clients who do not yet have Salesforce and large organizations that work on Salesforce and use its extensive capabilities ☁.

We also provide body and team leasing services in IT, providing specialists in various fields.

Model: remote

Employment type: full-time

Responsibilities:

• Design and implement complex multi-agent workflows using LangChain, LangGraph, or n8n, including advanced agent routing and state management.

• Build and deploy advanced RAG (Retrieval-Augmented Generation) systems in production environments.

• Develop and integrate multimodal conversational pipelines, including TTS/STT (Text-to-Speech / Speech-to-Text) for asynchronous Human-AI interview flows.

• Architect robust Text-to-SQL pipelines, accurately mapping natural language to deterministic SQL queries and predefined business logic views.

• Ensure AI systems strictly route financial and operational logic to validated backend services (no probabilistic “guessing” of calculations).

• Design and implement continuous evaluation frameworks, including LLM-as-a-judge validation pipelines to monitor hallucination rates and response quality.

• Implement AI observability and monitoring using tools such as Langfuse (or similar), creating a continuous improvement feedback loop.

• Develop multilingual validation pipelines to evaluate RAG performance, intent classification accuracy, and transcription quality (including Italian medical/pharmaceutical terminology).

• Optimize AI system performance through prompt caching, context-window management, and intelligent fallback model configuration to ensure reliability and low latency.

• Define guardrails and testing strategies appropriate for GenAI systems beyond traditional QA approaches.

Required:

• Proven hands-on experience with LangChain, LangGraph, or n8n (multi-agent workflow orchestration).

• Strong experience designing and deploying RAG systems in production.

• Practical experience integrating TTS/STT pipelines into conversational AI systems.

• Advanced knowledge of SQL and deterministic business logic routing (Text-to-SQL systems).

• Experience with Langfuse or similar AI observability/monitoring tools.

• Solid understanding of prompt caching and context-window management strategies.

• Experience designing evaluation and validation pipelines for GenAI systems.

• Strong system-thinking mindset with focus on reliability, optimization, and scalable AI architecture.