Senior AI Engineer

WHAT YOU'LL DO

Lead two parallel tracks:

(1) optimizing our existing warehouse (indexes, partitioning, data marts) and (2) building production AI capabilities (RAG, pgvector semantic search, MCP agentic workflows).

You will collaborate cross-functionally to make strategic architectural decisions, build natural-language querying infrastructure, and continuously iterate to accelerate business-critical reporting.

Ultimately, you will thrive in an entrepreneurial environment, continuously learning and iterating to better serve our business needs.

WHAT YOU'LL DO

• PostgreSQL Optimization: Diagnose and optimize slow queries (pg_stat_statements, EXPLAIN ANALYZE, B-tree/BRIN/GIN indexing). Build and materialize pre-joined data marts via dbt and Airflow.
• AI & Retrieval Infrastructure: Operate pgvector for semantic search and hybrid retrieval. Architect RAG pipelines and safely expose data/tools to LLM agents via the Model Context Protocol (MCP) with robust prompt-injection defenses.
• LLM Integration & Evaluation: Integrate major LLM providers (Anthropic, OpenAI, open-weights) balancing cost, latency, and accuracy. Build empirical evaluation harnesses for AI outputs.
• Cross-Functional Leadership: Partner with analysts and stakeholders to design reliable AI/data solutions. Review SQL, dbt models, and prompts while mentoring the team on engineering best practices.
• Architecture & Governance: Own the roadmap for potential lakehouse migrations (e.g., Databricks, ClickHouse). Enforce rigorous data quality and AI output governance standards.

WHO YOU ARE

• Education: Bachelor’s in CS, Data Engineering, ML, or related field.
• Experience & PostgreSQL: 5+ years in data engineering or applied AI/ML, with 2+ years scaling PostgreSQL (100M+ rows/multi-TB). Deep expertise in SQL, Postgres internals, indexing trade-offs, and CONCURRENTLY-aware deployments.
• Modern Data Stack: Hands-on experience with dimensional modeling, dbt, Airflow, Git workflows, and CI/CD for data.
• Applied AI: Proven track record shipping production LLM features (RAG, vector DBs, tool-using agents) and integrating major APIs with strict prompt and rate-limit management. Working knowledge of (or ability to rapidly learn) MCP.
• Programming: Strong Python skills; comfortable reading/writing production code in TypeScript, Go, or Java.
• Communication & Professionalism: excellent writer for technical docs (RFCs, post-mortems) and strong collaborative problem-solver.

DESIRED SKILLS

• AI Toolchain: Experience with LangGraph, LlamaIndex, DSPy, Pydantic AI, and LLM evaluation frameworks (Ragas, Braintrust, custom harnesses).
• Advanced Data Tooling: Production experience with Databricks and streaming/CDC solutions (Debezium, Kafka, Fivetran).

COMPESATION

Compensation package starts at 85k EUR / year and includes paid vacation days.