Senior AI Engineer - Conversational Analytics Agent

About the role

Next-Link is looking for a Senior AI Engineer / Agent Developer to join its consulting team and work closely with a key client in the pharmaceutical sector. The role contributes to a strategic Data & Analytics initiative focused on building an enterprise-grade conversational analytics agent for governed clinical data, combining Generative AI, agentic systems, advanced analytics and data governance.

Key responsibilities

• Design and develop a conversational analytics agent covering natural-language understanding, text-to-SQL, query execution, answer synthesis, explainability and source citation.
• Implement LLM agent architectures and orchestration patterns such as ReAct, plan-and-execute or multi-agent workflows.
• Build guardrails for hallucination prevention, traceability, scope enforcement and responsible AI.
• Integrate the agent with Snowflake, semantic-layer assets, business glossaries and KPI definitions.
• Implement secure data-access controls including RBAC and row-level security.
• Define and maintain structured AI evaluation frameworks covering SQL correctness, answer faithfulness, regression and drift.
• Implement observability, tracing and performance monitoring for agent behaviour, latency, token usage and errors.
• Establish CI/CD and automated regression testing for prompts, agent configurations and semantic artefacts.
• Collaborate with Data, Platform, AI Governance and Clinical Analytics teams.

Required profile

• Fluent English.
• Strong experience building LLM-based applications, text-to-SQL or conversational analytics solutions.
• Experience in regulated or governance-first environments, including RBAC, auditability and responsible-AI controls.
• Experience in pharma, healthcare, life sciences or another regulated industry is highly valued.

Required skills

• Python and advanced SQL.
• LLM patterns: prompt engineering, function calling, tool use and agent orchestration.
• Snowflake and familiarity with Snowflake Cortex.
• LangGraph and similar agent frameworks.
• Evaluation tools such as Ragas or DeepEval, and observability platforms such as LangSmith, Weights & Biases or Arize.
• Strong software-engineering, testing and CI/CD practices.