Data Platform Engineer

- 3 to 5 years of experience in data platform, data engineering, or backend/distributed systems engineering, including ownership of production systems at TB-scale, or systems carrying stringent SLAs, multiple downstream consumers, or compliance requirements - Strong Python, plus production experience in a JVM language, with the ability to write production-quality code with proper error handling, logging, and testing. Scala is our primary pipeline language: if you don't have it yet, you'll need a real appetite to become strong in it - Hands-on depth in either distributed batch processing (Spark: DataFrames, Spark SQL, partitioning, shuffles, fault tolerance) or streaming (Kafka or Pub/Sub: producers, consumers, topics, partitions, stream processing patterns), with working familiarity of the other

The platform: The platform has strong foundations: Scala and Spark pipelines processing TB-scale data, dozens of managed data connectors, Airflow on Astronomer for orchestration, a mature dbt transformation layer, and BigQuery as the core analytical warehouse. You'll also have room to shape the platform's next chapter: modernising parts of the compute stack, and building out governance frameworks, data contracts, SLOs, observability, cost management, and data quality monitoring within your domain. What You'll Do:

- Familiarity with dbt, Looker, or semantic layer concepts, and with dbt orchestration through Astronomer Cosmos - Functional Scala patterns (cats-effect and the Typelevel ecosystem), or JVM tuning exposure such as garbage collection, memory management, and thread pools - Observability tooling such as Grafana, Datadog, or Cloud Monitoring

- Strong Python, plus production experience in a JVM language, with the ability to write production-quality code with proper error handling, logging, and testing. Scala is our primary pipeline language: if you don't have it yet, you'll need a real appetite to become strong in it - Hands-on depth in either distributed batch processing (Spark: DataFrames, Spark SQL, partitioning, shuffles, fault tolerance) or streaming (Kafka or Pub/Sub: producers, consumers, topics, partitions, stream processing patterns), with working familiarity of the other - Advanced SQL, including complex queries, window functions, CTEs, and query optimisation

- Hands-on depth in either distributed batch processing (Spark: DataFrames, Spark SQL, partitioning, shuffles, fault tolerance) or streaming (Kafka or Pub/Sub: producers, consumers, topics, partitions, stream processing patterns), with working familiarity of the other - Advanced SQL, including complex queries, window functions, CTEs, and query optimisation - Experience with workflow orchestration such as Airflow: scheduling, dependencies, retry logic, and workflow coordination

What You'll Do: - End-to-end domain ownership: Own platform components and a pipeline domain end-to-end, including design, implementation, reliability, performance, and cost. Design and build scalable ETL/ELT and streaming pipelines using Spark/Dataproc, Kafka, and Pub/Sub. Integrate diverse sources into BigQuery and other stores. Build modular, fault-tolerant components with graceful schema evolution. - Infrastructure & performance tuning: Configure and maintain core platform infrastructure as code, including Dataproc, Kafka, Airflow/Astronomer, BigQuery, and Cloud Storage. Tune performance and cost through query optimisation, partitioning, and resource configuration. Tune streaming configurations for throughput and reliability.

- End-to-end domain ownership: Own platform components and a pipeline domain end-to-end, including design, implementation, reliability, performance, and cost. Design and build scalable ETL/ELT and streaming pipelines using Spark/Dataproc, Kafka, and Pub/Sub. Integrate diverse sources into BigQuery and other stores. Build modular, fault-tolerant components with graceful schema evolution. - Infrastructure & performance tuning: Configure and maintain core platform infrastructure as code, including Dataproc, Kafka, Airflow/Astronomer, BigQuery, and Cloud Storage. Tune performance and cost through query optimisation, partitioning, and resource configuration. Tune streaming configurations for throughput and reliability. - Reliability & incident response: Own and track SLOs for freshness, success, and latency. Implement monitoring, alerting, and data quality checks. Participate in on-call as a Tier 2 responder, troubleshooting and resolving incidents independently.

- Experience with workflow orchestration such as Airflow: scheduling, dependencies, retry logic, and workflow coordination - Working fluency with a major cloud platform, ideally GCP (BigQuery, Dataproc, Cloud Storage, Pub/Sub) - Solid engineering and operational practice: Git, CI/CD, automated testing, code review, monitoring, alerting, and incident response including on-call

- Functional Scala patterns (cats-effect and the Typelevel ecosystem), or JVM tuning exposure such as garbage collection, memory management, and thread pools - Observability tooling such as Grafana, Datadog, or Cloud Monitoring - Experience with Dataflow (Apache Beam), SingleStore, or BigTable

Impact's Data Platform Engineering team is looking for a Data Platform Engineer ready to independently own production pipelines and platform components at TB-scale, delivering data reliably, securely, and cost-effectively as the business grows. You'll design and build scalable batch and streaming pipelines, own the reliability, performance, and SLOs of your domain, and participate fully in on-call as a Tier 2 responder. You'll work within the Data Analytics Group (DAG), partnering closely with Analytics Engineers, Data Analysts, Data Scientists, and Data Product Managers, though you won't own dbt models, metric definitions, or reporting. We operate with a platform-as-a-product mindset: the platform is an internal product with clear interfaces, paved roads, and strong developer experience, and success is measured by the productivity of the teams who depend on it. Our ideal candidate combines solid distributed systems engineering with operational maturity, clear communication, and a habit of mentoring those earlier in their careers. The role reports to the Team Lead, Data Platform Engineering, and is based in Cape Town, hybrid, with two days per week in office. The platform:

- Experience migrating legacy big-data platforms onto cloud-native alternatives - Familiarity with dbt, Looker, or semantic layer concepts, and with dbt orchestration through Astronomer Cosmos - Functional Scala patterns (cats-effect and the Typelevel ecosystem), or JVM tuning exposure such as garbage collection, memory management, and thread pools

Nice to Have (Advantageous) Requirements: - Infrastructure as code (Terraform or equivalent) and automated deployment - Experience migrating legacy big-data platforms onto cloud-native alternatives

- Advanced SQL, including complex queries, window functions, CTEs, and query optimisation - Experience with workflow orchestration such as Airflow: scheduling, dependencies, retry logic, and workflow coordination - Working fluency with a major cloud platform, ideally GCP (BigQuery, Dataproc, Cloud Storage, Pub/Sub)