Join Amaris Consulting as a Senior AI Agent Developer within a clinical-operations data environment. In this role, you will contribute to the design, development, evaluation, and operation of an enterprise-grade conversational analytics agent for governed clinical-operations data, supporting high standards of reliability, traceability, and responsible AI.

Your missions

- Design and build a conversational analytics agent covering natural-language query understanding, SQL generation, query execution, answer synthesis, explainability, and source citations.
- Implement guardrails for scope enforcement, hallucination detection, traceability, and responsible AI in high-stakes clinical contexts.
- Select and document the agent orchestration approach based on proof-of-concept evidence, including ReAct, plan-and-execute, or multi-agent patterns.
- Integrate model access through the Apollo gateway while following cross-region inference and model-routing conventions.
- Implement purpose-bound data access according to user function, including object-level, attribute-level, and row-level controls.
- Integrate semantic assets such as business glossaries, KPI definitions, semantic views, and ontology references into the agent architecture.
- Design and operate a structured evaluation framework covering SQL correctness, answer faithfulness, regression testing, confidence intervals, and drift detection.
- Implement observability, CI/CD processes, and automated regression gates for prompts, semantic-model artefacts, and agent configuration.

Your profile
- Experience building text-to-SQL or conversational analytics systems in production environments.
- Experience developing production-level applications in Python.
- Advanced SQL skills, including analytical queries, window functions, and performance optimisation.
- Experience implementing LLM agent patterns such as prompt engineering, function calling, tool use, ReAct,



or plan-and-execute.
- Experience with Snowflake, including data modelling, SQL, and basic Snowflake Cortex concepts.
- Experience implementing agent orchestration through LangGraph, Snowflake Intelligence, Cortex Analyst, or a comparable solution.
- Experience using structured evaluation frameworks such as Ragas or DeepEval, and observability tools such as LangSmith, Weights & Biases, Arize, or equivalent.
- Experience in regulated, governance-first environments involving RBAC, data classification, audit trails, purpose-bound data access, and responsible-AI guardrails.
- Academic background in computer science, data science, artificial intelligence, software development, or a related field.
- English fluent.

What we offer
- An international community bringing together 110+ different nationalities.
- An environment where trust has a central place: 70% of our key leaders started their careers at the first level of responsibilities.
- A robust training system with our internal Academy and 250+ available modules.
- A vibrant workplace that frequently gathers for internal events, including afterworks and team buildings.
- At Mantu, sustainability is part of everything we do. You’ll have the opportunity to turn your ideas into action and make a tangible impact. Every day, our teams bring our ESG commitments to life, from reducing our footprint to driving positive change within our communities. Through our WeCare Together program, you’ll be empowered to design and lead projects that create real social or environmental impact, with the company’s full support.

Amaris Consulting is proud to be an equal-opportunity workplace. We are committed to promoting diversity within the workforce and creating an inclusive working environment. For this purpose, we welcome applications from all qualified candidates regardless of gender, sexual orientation, race, ethnicity, beliefs, age, marital status, disability, or other characteristics.

📌 Generative AI Engineer (Barcelona)
🏢 Mantu
📍 Barcelona