Why Doktor.se

Doktor.se is Sweden's largest digital healthcare provider. We combine physical clinics with a digital platform serving millions of consultations per year across Sweden, Germany, and Norway. Our 25-person engineering team has startup autonomy inside a profitable, international organization with 1,000+ employees and over 1B SEK in revenue.

We run OpenAI and Anthropic models in production across patient and clinician workflows — and that footprint is growing fast. Inside the engineering team, AI is a working tool, not a buzzword. We use it daily, we expect you to as well, and we hire people who treat it as a force multiplier.

What you'll do

Our data platform works, but it was built to solve yesterday's problems. We have data scattered across Postgres databases, Amplitude, S3 file dumps, and a tangle of Python jobs that move it around. The reports get built and the business runs on them. But trust is brittle, reconciliation is too manual, and the foundations need work before we can scale to where we're going.

We need someone who can help us rebuild this — not maintain what we have, but reshape it. If you want a place where the warehouse is clean and the pipelines just run, this isn't it. If you want to redesign a data platform from foundations up, with real ownership and the leverage to make it good, here's what you'd own:

• Reshape the data foundations. Multi-tenancy, retention, data contracts with backend services. The boring-but-load-bearing work that makes everything downstream possible.

• Decide what the warehouse actually is. Self-hosted Postgres + dbt, managed Snowflake or BigQuery, something else — we have opinions but not commitments. You'll shape this.

• Unify three data sources that don't currently talk to each other: our operational databases, Amplitude product analytics, and a long tail of S3-based partner integrations.

• Build observability into the platform. Lineage, freshness, alerting that actually means something. We've been flying blind for too long.

• Own reporting that finance, operations, and the board rely on. Make those reports trustworthy by design, not by manual reconciliation.

• Work directly with backend, product, and finance teams. You're not a service desk — you're a peer who shapes how data gets produced upstream, not just consumed downstream.

• Use AI aggressively. We try to avoid manually writing code that an AI can write for us. We expect the same discipline from you. Pipeline boilerplate, schema generation, transformation logic, exploratory analysis — if it can be AI-assisted, it should be.

Some weeks you'll redesign the warehouse. Some weeks you'll pair with backend engineers on a data contract. Some weeks you'll dig through messy partner exports to make finance's invoicing reliable. The common thread: you make data trustworthy, and you make the people who depend on it faster.

You'll be the only data engineer for the first few months while we hire. That's a lot of ownership and not a lot of cover. The right person sees that as opportunity.

What we're looking for

• You've designed and run data platforms in production. Not just written DAGs — you've thought about what to build vs buy, where transformations belong, how to keep schemas honest as products evolve.

• You're fluent in Python and SQL, and you have strong opinions about how Postgres should be used as a warehouse (and when it shouldn't).

• You've used dbt or equivalent in a real environment. You understand modelling — staging, marts, tests — not just transformation as a task.

• You have working knowledge of one or more managed warehouses (Snowflake, BigQuery, Redshift). You don't need to be an expert. You do need to be able to evaluate them honestly against self-hosted options.

• You're comfortable with orchestration tools — Airflow, Dagster, Prefect, whichever. You know what an orchestrator is for and what it's not.

• You think about data contracts as a product problem, not just a pipeline problem. You can have the conversation with backend engineers about why purging data after 14 days breaks reporting, and propose better patterns.

• You use AI in your daily work and you can talk about how. If your answer is "I haven't really tried it yet," this isn't the role.

• You have opinions — and you can change them when someone makes a better argument.

• Healthcare data experience is a plus. GDPR fluency is expected. Patientdatalagen familiarity is a bonus.

• You have a relevant university degree or equivalent experience.

Tech stack

• Sources: Postgres (operational), Amplitude (events), S3 (integrations and dumps)

• Languages: Python, SQL

• Cloud: AWS, multi-account

• Warehouse: Postgres today, open question for tomorrow

• Transformation: likely dbt going forward

• Orchestration: open question — you'll shape this

• AI: Claude and OpenAI in our daily workflow

What we offer

• Real influence. The data platform is being rebuilt and you're at the centre of it.

• Healthcare context. Patient data, regulatory compliance, real consequences for getting it wrong. The reports you build affect financial decisions and clinical operations.

• Small team, direct impact. 25 engineers — your work is visible, not lost in a machine.

• Autonomy. No micromanagement, no endless approval chains.

• AI as a working tool, not a side project.

• Collective agreement, 30 days vacation, wellness allowance, personal development budget.

Practical details

• Start: As soon as we find the right person. We'll accommodate your notice period.

• Employment: Permanent, full-time. Six month probation period.

• Location: On-site, Monday–Thursday at Sveavägen 63, Stockholm. Fridays are flexible.

• Language: You'll be working with our entire org so Swedish is a strong plus, but English is our working language within tech.

• Work permit: You must already have the right to work in Sweden.

We handle recruitment internally — no agencies, no consultants, no external recruiters.