Data Platform Engineer

Doghouse Recruitment; Amsterdam; On-site; EUR 90,000 - 125,000; Full time

Job summary

Doghouse Recruitment represents a growing trading firm where speed is the product. You will build a shared real-time event-streaming platform to replace bespoke pipelines, enabling data to move across the business with millisecond latency.

As the first hire on a brand-new data platform team, you design and operate the backbone using Kafka/Redpanda, Debezium, Terraform, ArgoCD and Kubernetes, with ownership and scalability at the core.

Qualifications

• Production experience with Kafka or Redpanda at meaningful scale.
• A track record of reasoning about latency, throughput, and reliability trade-offs in real-time or high-volume environments.
• Comfort with change-data-capture tooling and streaming/ETL patterns.
• Experience with infrastructure-as-code and GitOps (Terraform, ArgoCD, Kubernetes).

Responsibilities

• Designing and operating an event-streaming platform (Kafka or Redpanda) as the backbone for how data moves across the business.
• Building change-data-capture pipelines (e.g. Debezium) that pull data out of existing systems with minimal disruption.
• Making deliberate latency trade-offs and designing tiered delivery for fast paths.
• Running the platform on IaC and GitOps (Terraform, ArgoCD, Kubernetes) so it is reproducible and scalable.

Skills

Kafka/Redpanda scale

Latency optimization

Change-data-capture tooling

Infra as code & GitOps

Streaming/ETL patterns

Tools

Kafka

Redpanda

Debezium

Terraform

ArgoCD

Kubernetes

Job description

Our client is a growing trading firm where speed is the product. Every system from trade execution to the tools traders use during market hours, depends on data moving fast, reliably, and without breaking when volume spikes.

Today, every engineering team runs its own point-to-point data flow, tuned by hand. That doesn't scale. As the first hire on a brand-new data platform team, you'll build the shared foundation the rest of the organisation moves onto: a real-time event-streaming platform that replaces one-off pipelines with a single, well-understood standard.

What you'll actually work on:

• Designing and operating an event-streaming platform (Kafka or Redpanda) as the backbone for how data moves across the business.
• Building change-data-capture pipelines (e.g. Debezium) that pull data out of existing systems, including latency-sensitive trading systems, without disrupting them.
• Making deliberate latency trade-offs: most flows run comfortably at millisecond-scale, but you'll design a tiered delivery model so the handful of paths that genuinely need to be faster can be optimised separately, rather than over-engineering everything from day one.
• Running the platform on solid infrastructure-as-code and GitOps (Terraform, ArgoCD, Kubernetes) so it's reproducible and stays that way as usage grows.

You bring:

• Production experience with Kafka or Redpanda at meaningful scale.
• A track record of reasoning about latency, throughput, and reliability trade-offs in real-time or high-volume environments.
• Comfort with change-data-capture tooling and streaming/ETL patterns.
• Experience with infrastructure-as-code and GitOps (Terraform, ArgoCD, Kubernetes).

There's no platform yet, so you're not maintaining someone else's decisions, you're making them. Real ownership, real latency problems, and a green field to build the reference architecture the way you think it should be done.