Data Engineer – AI-Driven Data Platform (m/f/d)

Mercury Media Technology GmbH & Co.KG; Deutschland; Hybrid; EUR 90,000 - 120,000; Full time

Benefits offered by this job

Remote work within Germany

Office in Hamburg

AI coding agents access

Deutschlandticket

Hansefit

Job bike leasing

Company pension

Job summary

Mercury Media Technology GmbH & Co. KG is seeking a data-focused engineer to co-own the Data area of our media-planning platform. You will build and run data pipelines from advertising APIs to the analytics layer that powers media plans, dashboards, and billing-grade reports.

You will model and normalize data with dbt and SQL, ensure data quality with tests and checks, and extend our lakehouse platform. Remote work within Germany is supported, with a Hamburg office option.

Qualifications

• You have solid experience building and running data pipelines in production, from third-party APIs to curated, reportable data.
• You have very good knowledge of Python, especially for integrations against third-party HTTP APIs (pagination, rate limits, auth and token refresh, error semantics)
• You have very good knowledge of SQL and solid experience with dbt (incremental models, tests, and pitfalls like duplicates)
• You have good knowledge of PostgreSQL
• You have hands‑on experience with AI coding agents such as Claude Code, Codex, or Cursor
• You communicate clearly, especially in writing
• You care about data quality, tests, and architecture
• You enjoy talking to users and turning vague findings into concrete writeups
• You are fluent in English and/or German

Responsibilities

• You co-own the Data area of our platform with a small team, delivering end-to-end data pipelines from API to analytics.
• You turn customer needs into direction and shape how data is modeled and consumed.
• You maintain data connectors to platforms like Meta, Google Ads, TikTok, and others, handling endpoint changes.
• You model and normalize data with dbt and SQL, ensure data quality, and manage the lakehouse.
• You keep nightly pipelines reliable and drive root-cause analysis of incidents.

Job description

Your role

• You will co-own the Data area of our media-planning platform together with a small team of engineers and client relations — end to end, from the APIs of the advertising platforms to the analytics layer that powers media plans, dashboards and billing-grade reports
• AI agents write most of our code. You decide what gets built and how it fits together, turning customer needs and business outcomes into clear direction
• You will look after our data connectors to platforms such as Meta, TikTok, Google Ads/DV360, Amazon DSP, The Trade Desk, Adform, Pinterest and LinkedIn — and keep client data flowing when providers deprecate endpoints, rename fields, or silently change limits
• You will model and normalize data with dbt and SQL, and treat data quality as a feature: deduplication, freshness checks, and reconciling our numbers against what the platforms themselves report
• You will keep nightly client pipelines reliable and drive incidents to root cause — distinguishing provider outages, client-side setup issues, and real defects, and writing the findings down
• You will extend our multi-tenant semantic layer with new metrics and row-level security, and help build out our lakehouse platform
• You will work closely with users, client relations, and the other teams
• You will help shape how we work with AI. Our process changes as the tools improve, and we expect it to keep changing
• This role can be done remote within Germany or from our modern office in Hamburg

What you bring

• You have solid experience building and running data pipelines in production, from third-party APIs to curated, reportable data. You need to know what good data looks like to spot a silent problem in a pipeline you didn't build yourself
• You have very good knowledge of Python, especially for integrations against third-party HTTP APIs (pagination, rate limits, auth and token refresh, error semantics)
• You have very good knowledge of SQL and solid experience with dbt (incremental models, tests, and an instinct for pitfalls like duplicates, late-arriving data, and identifier mismatches)
• You have good knowledge of a relational database such as PostgreSQL
• You have hands‑on experience with AI coding agents such as Claude Code, Codex, or Cursor, and a realistic sense of where they do well and where they go wrong
• You communicate clearly, especially in writing
• You care about data quality, tests, and architecture, and you're comfortable saying no to a change that works but doesn't fit
• You enjoy talking to users and turning a vague "the numbers look wrong" into a concrete, written finding
• You are fluent in English and/or German. Our company language is English, so German is not required
• Knowledge of Snowflake, Trino, Iceberg, columnar data storage likeClickHouse, Airbyte, or Cube.js is a plus, as is Kubernetes/Terraform exposure, Elixir or Ruby, or experience in media or advertising
• Work permission and residence in Germany or willingness to relocate to Germany

You don't tick every box? Apply anyway. We'd rather meet an ambitious, sharp mind who has never worked with Trino or Iceberg than someone who ticks every box but doesn't fit our team.

How we build software

Our platform is split into areas such as Media Planning, Reporting Center, Data, API and Media Mix Modelling. Each area is owned by a small, stable team of people who know it deeply, even though they don't type most of the code themselves.

• People set direction, agents implement. The hard part is no longer typing code. It's knowing what to build and keeping the platform coherent. That's where you spend your time.
• Leadership set direction, teams set the order. Leadership describes outcomes: what success looks like, not how to get there. Your team decides what to work on and in which order.
• Few meetings. Each team keeps its own backlog and meets when it needs to. No long planning documents, no multi-week development cycles, no formal team lead. The team makes decisions together.
• Small things just get done. A customer‑reported bug with a small, contained fix doesn't wait for a meeting.
• How hard it is to undo decides who decides. Decisions that are easy to undo are made by the team. For the ones that are hard or expensive to undo, we bring in leadership before we commit.
• Ownership means looking after an area, not guarding it. Any team can change any area, and lets the owning team know.

Nobody has fully figured out how to run engineering with AI agents yet, including us. We keep improving our process as we learn, and you'll help shape it.

How we work together

• Remote, but not apart. We are a remote (within Germany) company but still value face to face meetings such as our monthly company meeting, parties, game nights or cultural events. Every employee is welcome to work from our beautiful office in Hamburg. However, we see it as everyone's own responsibility to decide how often they want to visit our office.
• Fast, written communication. As knowledge transfer is even more important in a remote environment, efficient and fast communication is a key factor for us. For this purpose we use Slack as our main communication tool.
• Decisions as a team. Teamwork is an essential part of our culture. Every area team makes decisions together, and teams can call an ad‑hoc meeting or bring in people from other teams whenever that helps. We stay on an area long enough to know it really well, and move between areas over time to share knowledge and get fresh ideas.
• Flexibility for families. We highly value self‑responsibility and flexibility which is reflected in our flexible working times and independent task completion. As we are very family‑oriented and many of us have children, we know that a certain degree of work flexibility helps a lot combining private obligations with work life.
• Room to grow. At MMT, we appreciate passion and growth. This is why we like to help everyone to be the best version of themselves. For this reason, we have regular appraisal, feedback and development talks where needs can be placed. We offer many opportunities for growth such as visiting conferences, trade shows, tech‑events or other training options, and time to try out new AI tools and ways of working.

Who you'll work with

We see our diversity as a key factor of our success and continuing growth. Our team currently includes 9 different nationalities and 17 different languages, which is why we have chosen English as our common corporate language. This diversity of cultural and professional backgrounds allows us to bring different perspectives and innovative solutions to the table. Each team member brings a unique set of experiences and ideas that enrich the way we work and help us respond creatively to challenges.

As a tech company, we know that our industry tends to be male‑dominated. Therefore, it is important for us to make working at MMT attractive to all genders and to meet their different needs. At MMT, everyone should have the same opportunities, regardless of gender, because we are convinced: Diversity leads to better ideas, innovations and ultimately to our joint success.

Our tech stack

• Connectors: Python (Airbyte‑based)
• Transformation: dbt, SQL
• Storage and query: PostgreSQL, Snowflake, Trino, Apache Iceberg (lakehouse)
• Semantic layer: Cube.js
• Orchestration and infrastructure: GitLab CI/CD as orchestrator, Docker, Kubernetes, Helm, Terraform
• Services at the edges of the data platform: Ruby on Rails, Elixir
• AI: Claude Code, with shared skills and plugins that we maintain as a team

What we offer

• Small and international teams where you can actually make a difference
• The opportunity to offer modern and leading solutions in an exciting industry
• A modern workplace in the center of Hamburg
• Flexible working hours with remote work up to 100%
• A laptop running Linux or Mac
• Company-paid access to leading AI coding agents
• A subsidized ticket for public transportation (Deutschlandticket)
• Subsidized gym membership (Hansefit)
• Job bike leasing
• Company pension scheme
• Space and time for personal growth and learning new things

You have kids, live far away, or prefer a non‑normative lifestyle – we'll make it work!

Who we are

MMT builds software that helps media agencies and advertisers plan, run and measure their campaigns, from planning and budgets to reporting. Founded in 2016 as a spin‑off of one of Germany's largest independent media agencies, we are now around 30 people, most of us in development.

Want to know more? Visit our careers page or read what employees say about us on kununu.

Non-discrimination policy

Mercury Media Technology GmbH & Co. KG does not and shall not discriminate on the basis of race, color, religion (creed), gender, gender expression, age, national origin (ancestry), disability, marital status, or sexual orientation, in any of its activities or operations. These activities include but are not limited to, the hiring of staff, selection of vendors, and provision of services. We are committed to providing an inclusive and welcoming environment for all members of our staff, clients, subcontractors, and vendors.