Data Analyst (Analytics Engineering)

OpenHealth Technologies; Polska; Remote; PLN 120,000 - 180,000; Full time

Job summary

Open Health Technologies is seeking a Data Analyst (Analytics Engineering) to own transformation models and analytic datasets on Dataform for BigQuery, shaping analytics that feed product decisions.

You will work with founders, maintain data quality, and build models that enable non-technical colleagues to answer questions directly, with emphasis on correctness and trust in the numbers.

Qualifications

• Analytical judgement: knowing which analysis is worth running and why, and what a result means for the business.
• BigQuery production fluency, cost awareness, and performant queries on large datasets.
• Dataform or dbt ownership: designed, maintained, refactored transformation models.
• Hands-on and independent: able to define and drive analyses without a handed-down spec.
• Fluent English communication: present findings to non-technical audiences and align founders.

Responsibilities

• Own analytics topics end-to-end, from a loosely defined business question to a delivered analysis.
• Build and maintain transformation models and analytical datasets in Dataform on BigQuery.
• Model data to enable self-serve answers rather than gating every question.
• Guard data quality and trust in the numbers; ensure definitions are consistent.
• Set direction in analytics and articulate what to build next to the founders.

Skills

BigQuery

Dataform

dbt

Analytical judgement

English fluency

Hands-on

Python

Pandas

Stakeholder comms

Data quality

Tools

SQL

Job description

Engagement: Full-time, embedded with our team (B2B / contract)

Location: Remote

Start: ASAP. We prioritise fit over speed.

Language: Fluent English is required, it is our company language. Any further language is a plus, German, Portuguese (Brazilian) and Spanish especially.

Commercial: Initial period 2 month, extension expected on mutual fit.

Context

Open Health Technologies is a healthcare data platform. We ingest heterogeneous lab and health data from many sources, harmonise it, and enrich it with context. It lands in a single data warehouse on GCP / BigQuery, with logical separation between tenants.On top of that sits OHA Cortex, our data intelligence layer: the analytics and insights product built on the harmonised and contextually enriched data. This role is for Cortex.

Where we actually are

The engineering groundwork is done. The warehouse already ingests from application-database CDC, webhooks (web tracking) and pub/sub. Transformation runs on Dataform. Analytical results are written back into the application, where they reach users, over a path we have already built.So this is not a greenfield build and it is not a rescue. The plumbing works. What is missing is someone to own what sits on top of it.

Scope of this role

This role owns the analytics layer: the transformation models, the analytical datasets, and the answers people act on.It does not cover ingestion pipelines or platform infrastructure. That work is done, and further work of that kind sits with a separate integration role we staff independently. Please do not send data-engineering profiles against this brief.The title reads “Data Analyst (Analytics Engineering)” deliberately. We need someone who owns transformation models and answers business questions. A pure report-builder is under-scoped for this. A pipeline engineer who does not want to face stakeholders is mis-scoped.

What makes this role unusual

Analytics output here does not stop at a dashboard. Models are written back into the product and reach users. That raises the bar on correctness, and it means the person in this seat is shaping the product rather than reporting on it.It also sets the priority order. Trust comes first. Data quality is the foundation of everything Cortex claims, and we would rather have fewer numbers we can stand behind than more numbers we cannot. We expect this person to be the one who cares most about that in the building.

What the role is

• Own analytics topics end-to-end, from a loosely defined business or client question to a delivered, well-communicated analysis.
• Build and maintain transformation models and analytical datasets in Dataform on BigQuery.
• Model data so that others can answer their own questions, rather than becoming a queue that every question has to pass through.
• Guard data quality and trust in the numbers. Correctness, definitions people agree on, and the discipline to flag when a result does not hold up.
• Set direction in the domain. Over time we want this person telling us what to build next in analytics and why, not waiting to be told. That is the growth path we care about.

How we work together

This is an ownership role, not a solo one. The person in this seat works in regular direct dialogue with the founders (CEO, CPO, CTO), and we expect alignment to be maintained continuously rather than reported at the end.We are looking for someone who forms their own view, argues it clearly, and then commits to where we land together. Neither a lone hero who disappears and returns with a finished opinion, nor someone who waits for a spec.

Must-haves

• Technical skill is necessary here but it is not what the role turns on. We have met strong engineers who could build any model asked of them and had no view on which model was worth building. We are hiring for both halves.Analytical judgement. Knowing which analysis is worth running and when, which question behind the question is the real one, and what a result actually means for the business. This is the half that cannot be taught quickly, and the half we will probe hardest.
• BigQuery production fluency, not SQL literacy. Real production experience: performant queries on large datasets, partitioning and clustering, cost awareness. We mean demonstrated ownership of a warehouse in production, not picking up the dialect in a couple of days on top of another cloud.
• Dataform, or dbt, which we treat as directly transferable. What matters is having owned transformation models: designed, maintained, refactored, and been responsible when they broke. Consuming someone else’s models does not count.
• Hands-on and independent. A track record of defining and driving analyses without being handed a spec. This seat is hands-on for the foreseeable future.
• Fluent, efficient communication in English. Able to present findings to non-technical audiences, defend analytical choices under questioning, and keep busy founders aligned without long documents.
• Rigour about data quality. Someone who checks before they publish and says so when a number is shaky.

Nice-to-haves

• Healthcare or lab data exposure (LOINC, UCUM, FHIR, or similar standards)
• Self-serve BI experience, modelling data so non-technical colleagues can explore it themselves
• Comfortable dropping into Python (pandas, notebooks) when SQL runs out. Our current work is overwhelmingly SQL and Dataform, so this is a bonus rather than a bar.
• Broader GCP ecosystem experience
• Small-team or startup experience, comfortable with evolving requirements and short decision paths

Seniority, and an honest word on leadership

Mid-level to senior individual contributor.

To be plain about it: there is no analytics team to lead today. For the foreseeable future this is a hands-on seat, and the person in it will be the first and only analytics person. Leading a team later is a genuine possibility and we would rather grow it from inside than hire it in, but we are not going to dress it up as a near-term promise.We stay small on purpose: at most around 3 hires over the next six to twelve months, chosen deliberately. This is an ownership role before it is a leadership role, and we are looking for someone who finds that appealing rather than disappointing.

The first six months

• Get up to speed fast. The setup, the infrastructure, the team, and above all the product vision. We will support this actively, and we expect someone who pulls rather than waits.
• Establish the dialogue with the founders. Regular, direct contact with CEO, CPO and CTO so that priorities stay aligned as they move.
• Hands-on work that grows the product. Real analytics shipping into Cortex, not a three-month audit phase.

By the end of it, we want someone who knows the domain well enough to start telling us what comes next.