You will join as the anchor data engineering hire at a scale-up operating inside a larger group, with the executive team backing the build.
This is an individual contributor role with reach across the whole organisation, so your influence rests on technical credibility rather than team size.
The work is greenfield and high stakes: the feature store and the agentic workflows built on it will carry design decisions that last for years.
You would partner closely with the Head of Data Engineering and the Head of ML.
The company runs on a modern stack and a lean, AI-native team.
The team's core value is extreme ownership.
Responsibilities
Design and own the feature store architecture that credit decisions will rely on.
Define schemas, point-in-time correctness and freshness targets, aiming for designs that are dependable and inexpensive to run.
Set the pipeline patterns, modelling standards and lineage approach that other engineers adopt as reference implementations.
Lead the most complex and risky data initiatives, including feature reuse across regulated entities.
Design agent loops covering context, tools, evals and guardrails, so agentic engineering is reliable, observable and cost-controlled.
Own the architecture and internals of agentic harnesses and loop or graph workflows tied to the feature store, including feature engineering, point-in-time validation, and drift and data quality observability.
Establish reusable reference patterns for that agent work.
Shape the quality and speed of revenue-generating decisions through your architecture.
Requirements
You enjoy working with data and building systems around it.
You build hands-on and use AI tools daily in your work.
8+ years of data engineering experience at scale.
5+ years of experience with Spark, Databricks or comparable lakehouse architecture.
7+ years of Python, plus expert SQL and data modelling skills and experience engineering for performance and cost.
3+ years using agentic coding tools and loop engineering, including context and tool design, evals and guardrails.
Interest in advanced agentic harness, loop and graph engineering, with solid knowledge of LLM behaviour and trade-offs.
Staff-level engineer who sets standards and mentors through influence rather than authority.
Strong in data quality, observability, lineage and GDPR-aware governance.
Fluent English, B2+.
Bonus: feature store or ML data serving in production.
Bonus: payments, fintech or lending data experience.
Bonus: streaming with Kafka or Structured Streaming, and CDC.
Bonus: dbt governance experience.
Bonus: awareness of CCD2 and DORA.
Conditions
Hybrid work, with 1 day a week in the Warsaw or Kraków office.