1 day ago
Madrid, SpainStaff+
Responsibilities
- Own and evolve the technical architecture for DevHub’s AI Gateway, agentic services, retrieval/context pipelines, and platform-AI integration points.
- Create and maintain architecture decision records and RFCs, and serve as the architectural authority for patterns, technology choices, and build-versus-buy decisions.
- Design reusable AI platform components including context engines, guardrail and evaluation libraries, MCP/tool layers, scaffolder templates, SDKs, APIs, and versioned interfaces.
- Design MCP, SDK, and API surfaces that are read-first, RBAC-aware, and tenant-aware for internal and external AI clients.
- Define evaluation frameworks using eval harnesses, golden tests, retrieval-quality metrics, grounding controls, and hallucination controls.
- Shape MLOps and LLMOps practices for prompt and version management, model adaptation, and production delivery of AI artifacts.
- Architect multi-tenancy, governance, observability, distributed tracing, per-product cost visibility, RBAC, auditability, and AI compliance controls.
- Mentor AI engineers, conduct architecture reviews, pair on complex problems, and raise the team’s technical standards.
- Represent AI architecture in discussions with Enterprise Architecture, Product Family leadership, Databricks, Microsoft, and other external partners.
Requirements
- BSc or MSc in Computer Science, Software Engineering, AI, or a related field.
- At least 7 years of software or platform engineering experience, including 2–3 years in technical leadership or architecture roles and recent generative AI or AI-driven systems experience.
- Expert Python expertise and production AI/LLM architecture experience, including FastAPI, Pydantic, and asynchronous programming.
- Deep experience designing production-scale LLM systems using RAG architectures, agentic patterns, context engineering, and autonomous agents appropriately.
- Strong system design experience for distributed, multi-tenant systems, including API boundaries, data flows, resilience, ADRs, RFCs, and architecture diagrams.
- Experience designing evaluation frameworks for LLM/RAG systems, including eval harnesses, golden tests, retrieval-quality metrics, grounding, and hallucination controls.
- Experience designing reusable self-service platform components, templates, SDKs, and golden paths across teams.
- Deep Azure experience, Kubernetes or AKS at scale, Docker, and infrastructure as code using HCL or Terraform.
- Experience with PostgreSQL, vector stores such as Azure AI Search, and state management in multi-tenant platforms.
- Knowledge of OpenAI, Anthropic, Mistral, Azure AI Foundry, Databricks, and MCP.
- Strong understanding of authentication, authorization, RBAC, tenant isolation, guardrails, auditability, OWASP LLM Top 10, and EU AI Act implications.
- Experience mentoring engineers, conducting architecture reviews, setting technical standards, and communicating technical decisions clearly.
- Preferred qualifications include Databricks Unity Catalog, TypeScript, Grafana, MLflow, FinOps for AI, A2A, agentic development standards, RabbitMQ, Azure Event Grid, and enterprise architecture or vendor-team experience.
Benefits
- Secure work environment with a focus on health, safety, and wellbeing.
- Flexible work schedule and home-office options.
- Learning and development opportunities.
- 25 holiday days per year plus 5 additional readjustment days.
- Collaborative, trustful, innovative, international, and global-project work environment.
- Relocation assistance to Madrid.
Tech Stack
Categories
About BASF SE
BASF SE is a public chemical manufacturer that produces basic chemicals, advanced materials, surface technologies, and agricultural solutions for customers in industries such as automotive, construction, and consumer goods, as well as farmers. It sells globally through a portfolio of business segments and R&D-driven product lines. Founded in 1865 and headquartered in Ludwigshafen, Germany, BASF is listed on the Frankfurt Stock Exchange.