Senior Data Engineer

Engineering · Amsterdam (hybrid)

Inner Circle is a dating app for people who are done wasting time. Every member is screened thoroughly, so the profiles you see are people worth dating.

All of that, the screening, the matches, the marketing, the subscriptions, runs on data. Every AI model and feature that we ship next is built on the data you own.

Goal of the Role

You build and run the systems that move our data; from the product database, app events, attribution and payment providers into Snowflake - the central analytical database that streams data to various destinations. Everything the company decides sits on top of what you build.

You'll be part of our engineering team, reporting to the Head of Engineering. But you set the technical direction for data: architecture, tooling, standards and roadmap come from you, backed by reasoning your manager can follow and trust.

We're building toward being an AI-native company, and data is the constraint on how far that goes. Claude Code is part of how we work day to day. Engineers build with it, and we're moving toward agentic workflows for things like error triage and bug fixing. Every AI use case we take on, whether that's smarter screening, matching, or internal tooling, ends up depending on the data being correct, timely and well-modelled. That's the platform you own.

Tech stack: Snowflake, DBT, Dagster, DataBricks, Python, AWS, BigQuery, MySQL, GitHub, ETL/ELT, Tableau, Claude Code, DataDog

Key Result Areas

1. Own the direction of the data platform

This is yours to shape. You form a view of where the platform needs to go, make the call, and bring the reasoning rather than the question. Tooling and architecture decisions come with the tradeoffs; stated cost, complexity, who maintains it, what happens when it breaks. Your roadmap separates the urgent from the structural, and the structural work actually happens.

2. Build and run reliable pipelines

You own how data moves, end to end. The best pipelines are boring ones: they run, and when they don't, you know before anyone else does. Failures are caught by monitoring, not by someone asking why a dashboard is empty. New sources and markets get integrated properly.

3. Engineer for correctness

Correctness is designed in, not checked afterwards. Tests sit at the right layers and failures are actionable instead of noise. Models are incremental and idempotent, so the same run gives the same result. The messy parts; deduplication, late-arriving data, timezones, attribution windows, are handled deliberately.

4. Set the engineering standard for the data stack

You define how data gets built here, and the standard you set outlasts you. Everything is version-controlled, reviewed and covered by CI. Nothing reaches production by hand. Code is readable enough that someone else can pick it up cold, because at some point someone will. Warehouse spend is watched, and access is managed as code.

5. Be the data counterpart the rest of the company can rely on

Data isn't a reporting function here. Screening, matching, marketing, subscriptions and increasingly our AI tooling all run on what you build, which means most of the company touches your models whether they know it or not. You're close to the business, upstream and downstream, so your work lands visibly and fast.