Staff Engineer — Data Platform

Описание

As a Staff Engineer - Data Platform on the infrastructure department, you’ll play a vital role in building the core data platform that powers Taboola’s reporting, billing, ML training, experimentation, revenue optimization, and business decision-making.

Responsibilities

• Shape the technical direction of the Data Platform group: the architecture, the long-term roadmap, and the bets on emerging tech (Iceberg, StarRocks, GPU Spark, dbt, AI tooling)
• Mentor and grow engineers across the group: lead design reviews, pair on hard problems, review code with care, and help strong engineers grow into technical leaders
• Raise the engineering bar group-wide: set standards for design, code quality, testing, observability, and operational excellence
• Own the hardest problems on our stack: Spark performance and cost across 60,000+ jobs/day, Iceberg table management at Taboola scale, high-throughput Kafka ingestion, and low-latency analytics on StarRocks
• Evolve our Spark SQL platform and the core framework code used across R&D: make the call on what gets built, refactored, or retired
• Drive the evolution of our lakehouse, with Iceberg as the backbone: from ingestion through storage, compaction, and serving
• Be the go-to expert for the production health of the main Taboola data pipeline: the system behind reports, ML model training, billing, and other business-critical processes
• Lead deep-dives and postmortems on the toughest incidents
• Partner with team leaders, group managers, and fellow principal engineers to shape how the platform evolves across R&D, balancing the needs of algo, product, finance, and infrastructure stakeholders
• Stay deeply hands-on: write production code, build prototypes, and prove out new technology before the group commits to it
• Represent the group internally and externally: in architecture forums, hiring, open-source communities, and conference talks
• Directly shape how billions of people discover the internet

Requirements

• 10+ years of hands-on software engineering, with deep expertise in Java or an equivalent JVM/object-oriented language
• Deep, hands-on expertise in Apache Spark: internals, Spark SQL and the Catalyst optimizer, and performance and cost tuning at scale
• Production experience with Apache Iceberg (or another open table format such as Delta Lake or Hudi): table design, partitioning, compaction, and metadata/catalog management
• Deep expertise in Apache Kafka: internals, partitioning and delivery semantics, and operating high-throughput pipelines in production
• Hands-on experience with StarRocks or a similar MPP/real-time analytics engine (Druid, ClickHouse, Trino)
• Lead by example with AI: introduce agent-based engineering workflows, measure their impact, and help the group adopt them
• A track record of designing, building, and owning large-scale distributed data systems end to end: from architecture to production reliability
• Proven experience mentoring engineers and growing technical leaders: through design reviews, code reviews, pairing, and knowledge sharing
• Ability to influence without authority and drive alignment across teams, managers, and stakeholders
• Highly proficient with Linux and Kubernetes in production
• B.Sc. in Computer Science or equivalent experience

Nice to have

• Contributions to open-source data projects (Spark, Iceberg, Kafka, StarRocks, or similar)
• Conference talks, blog posts, or other public technical writing
• Experience running Spark on GPU (RAPIDS) or other query acceleration techniques
• Experience with stream processing (Spark Structured Streaming, Kafka Streams, Flink)
• Experience with dbt and/or Airflow
• Knowledge of Ansible, Puppet, or similar configuration systems
• A track record of leveraging AI agents to accelerate engineering work
• Familiarity with how ML training pipelines consume data, to partner effectively with algo teams

We offer

• Competitive benefits and well-being programs tailored to the region, including comprehensive health benefits, a fully stocked kitchen, and location-specific perks
• Hybrid work schedule with 3 days in-office with an option to come in more often if desired
• Work with some of the biggest names: publisher partners include Yahoo, Conde Nast, Fox Sports, NBCU, ESPN, CBS, and E! Online; advertiser clients include Wells Fargo, Honda, Pinterest, and Expedia