CareerPlanGet AI match score →

Staff Backline Engineer - Data & AI

United States💼 Full-time💰 $170–$170🗓 2026-04-09 → 2026-07-31

Core

Bridge between Frontline Support and Engineering to resolve complex technical issues, perform root cause analysis, and drive platform improvements for the Data and AI ecosystem.

Role type

Staff Backline Engineer (Data & AI)

Builds

Reproduction frameworks, automated workflows, and AI-driven diagnostic tools to enhance troubleshooting efficiency and platform supportability.

Domain

Data & AI infrastructure, distributed systems, Spark, Delta Lake, and large-scale ML.

Deliverable

production ML models | infrastructure

Required skills

Deep expertise in Spark, Delta Lake, or Hive; code-level root cause analysis and profiling in Java, Scala, or Python; distributed system internals; large-scale ML and generative AI systems; LLM-based applications and agent-driven workflows; model training, evaluation, and deployment; managing the ML lifecycle; diagnosing and optimizing distributed ML workloads.

Preferred skills

Experience managing customers and technical stakeholders; mentoring other engineers; influencing product roadmap.

Technologies

Spark, Delta Lake, Hive, Python, SQL, Scala, Java, LLMs, MLflow

Responsibilities

Conduct deep-dive forensics into Spark core internals and the Databricks ecosystem to resolve high-priority architectural failures; Perform advanced code-level analysis and resource profiling to identify and mitigate systemic root causes; Optimise architectural performance across the Data and AI stack by refining execution parameters; Analyse global issue trends and patterns to partner with Product Engineering and drive initiatives that enhance long-term supportability; Develop reproduction frameworks, automated workflows, and AI-driven diagnostic tools to empower and scale the broader organisation.

Seniority

Staff, hands-on IC with strategic influence

Sourced via greenhouse · Listed on CareerPlan, which tracks 70,000+ jobs from 20+ sources.
Apply on Greenhouse ↗