Principal Data Platform Engineer (Swedish Ad Platform)

Our Customer is a Sweden-based AdTech company specializing in advanced self-serve advertising platforms that automate direct transactions between advertisers and major global publishers. Their solutions improve transparency and operational efficiency in digital advertising and are trusted by globally recognized brands including TripAdvisor, Bloomberg, The Washington Post, Opera, and Dow Jones. The company processes millions of advertising transactions worldwide and is actively investing in AI-driven data capabilities.

PROJECT

The project focuses on building a modern data platform from the ground up with an emphasis on immutable event streams, governed canonical models, semantic layers, and scalable analytics infrastructure. The platform will support analytical workloads, operational data products, AI applications, and future customer-facing data experiences.

As a Principal / Staff Data Platform Engineer, you will own key architectural decisions, define scalable engineering standards, and help deliver the first production-ready version of the platform in a high-scale AdTech environment.

Key Technologies: Apache Iceberg, Spark, Flink, Trino, AWS, Python, Scala, Java, CI/CD, Infrastructure as Code

Job Description

Lead the technical design and architecture of a modern enterprise-scale data platform

Define scalable data flows from Iceberg-based event storage into analytical and operational data products

Evaluate and select technologies for orchestration, transformation, querying, serving, and storage

Design reusable canonical entities and data models across advertising, campaigns, inventory, billing, and customer domains

Establish scalable engineering patterns for batch and near-real-time processing

Build reliable and observable data pipelines for high-volume AdTech workloads

Implement data quality, lineage, observability, and reconciliation capabilities

Collaborate with Platform Engineering teams to introduce CI-enforced data contracts and governance standards

Define tenant isolation, access control, and regional data boundary strategies

Establish standards for testing, deployment automation, schema evolution, and versioning

Optimize platform performance, scalability, and infrastructure costs

Mentor engineers and contribute to engineering excellence across the team

Partner closely with leadership and cross-functional stakeholders

Qualifications

8+ years of experience designing and operating production-grade data platforms

Strong expertise in distributed data systems and high-volume event-driven architectures

Deep understanding of modern lakehouse architectures and Apache Iceberg

Advanced SQL skills and production experience with Python, Java, Scala, or similar languages

Experience with distributed processing and query technologies such as Spark, Flink, or Trino

Strong knowledge of data modeling, partitioning strategies, and performance optimization

Proven experience building batch and near-real-time data pipelines

Hands-on experience with AWS cloud infrastructure

Strong understanding of CI/CD pipelines, Infrastructure as Code, and production observability

Experience implementing data contracts, schema evolution, and data quality frameworks

Ability to make pragmatic architectural decisions in greenfield environments

Upper-Intermediate or higher English level

Strong communication and technical leadership skills

Will be a plus

Experience in AdTech or other high-volume event-processing domains

Experience designing multi-tenant SaaS data architectures

Familiarity with semantic layer technologies

Experience supporting both analytics and ML/AI workloads

Understanding of privacy regulations and data residency requirements

Additional information

What success looks like within your first six months:

The first production data foundation is running.

Clear architectural decisions have been made and documented.

Events entering the data platform have enforceable contracts.

Core canonical entities exist and are tested.

Data quality and lineage are observable.

Other engineers can contribute without needing to understand every implementation detail.