2 hours ago
Madrid, SpainMid Level
Responsibilities
- Build reusable AI platform components, including retrieval/context engines, guardrail and evaluation libraries, MCP/tool layers, templates, SDKs, and versioned APIs.
- Develop grounded assistants, copilots, and wizards for BASF’s portal, IDE, and Teams through shared APIs.
- Expose read-first, RBAC- and tenant-aware platform capabilities for human and agent clients.
- Create evaluation harnesses, golden tests, retrieval-quality metrics, and CI regression testing for LLM and RAG features.
- Improve prompt and version management, model adaptation, CI/CD, and the path from experimentation to production.
- Operate multi-tenant AI services with observability, distributed tracing, per-product cost tracking, FinOps visibility, and governance.
- Apply deterministic pipelines, structured outputs, human-in-the-loop controls, and gated agentic workflows where appropriate.
- Support security, safety, auditability, RBAC, tenant isolation, OWASP LLM Top 10 defenses, EU AI Act compliance, and data-residency requirements.
Requirements
- BSc or MSc in Computer Science, Software Engineering, AI, or a related field.
- At least 4 years of software engineering or platform development experience, with recent Generative AI and/or agentic systems experience.
- Strong Python skills and hands-on experience building LLM-powered applications.
- Practical familiarity with RAG, agentic patterns, context engineering, and deciding when autonomous agents are inappropriate.
- Experience designing evaluation harnesses, golden tests, retrieval-quality metrics, grounding controls, and hallucination controls for LLM/RAG systems.
- Backend and API development experience, preferably with FastAPI, Pydantic, and asynchronous Python services.
- Experience building reusable self-service platform components, SDKs, templates, golden paths, and production multi-tenant services.
- Working knowledge across frontend, backend, infrastructure, Azure, Docker, Kubernetes/AKS, and Terraform/HCL.
- Experience with PostgreSQL, vector stores such as Azure AI Search, Git, GitHub Actions, monitoring, observability, authentication, authorization, and RBAC.
- Familiarity with OpenAI, Anthropic, Mistral, Azure AI Foundry, Databricks, and the MCP standard.
- Preferred qualifications include Databricks/Unity Catalog, TypeScript, Grafana, MLflow, data pipelines, FinOps, RabbitMQ, Azure Event Grid, workflow engines, OpenTelemetry, distributed tracing, and EnvoyProxy experience.
- Familiarity with A2A, spec-driven agentic development, AGENTS.md, skill conventions, and the OWASP LLM Top 10 is beneficial.
Benefits
- Flexible work schedule and home-office options.
- Learning and development opportunities.
- 25 holiday days per year plus 5 additional readjustment days.
- Collaborative, trustful, innovative, and international work environment.
- Participation in global projects.
- Relocation assistance to Madrid.
- Secure work environment with emphasis on health, safety, and wellbeing.
Tech Stack
AzureDatabricksDockerFastAPIGitGitHub ActionsGrafanaKubernetesMLflowPostgreSQLPythonRabbitMQTerraformTypeScript
Categories
