Software Engineer – Data Flywheel (Member of Technical Staff)

About the role

Perplexity is building an LLM‑first search engine that serves tens of millions of users daily. The Answer Quality team needs a Software Engineer to create the data flywheel that powers fast, accurate, and actionable evaluations across the product.

Key responsibilities

• Design and implement systems and pipelines that let Search, Product, and other teams retrieve reliable evaluation verdicts without bottlenecks.
• Own the end‑to‑end "evals‑to‑product" loop, turning raw signals into durable datasets for company‑wide decision making.
• Build a robust simulator pipeline that replays user interactions in formats consumable by LLMs and VLMs, reflecting product changes in real time.
• Maintain data trust through monitoring, lineage, and quality checks so downstream consumers can rely on results implicitly.
• Collaborate in a small, high‑impact team where your work directly shapes Answer Quality measurement.

Required profile

• 3+ years of software engineering experience delivering production systems.
• Strong proficiency in Python and SQL, with production‑grade, maintainable code.
• Experience with big‑data ecosystems, distributed compute, and large‑scale storage.
• Solid fundamentals in data modeling, system design, and debugging distributed systems.
• Hands‑on experience with AWS and lake‑house platforms such as Databricks or Spark.
• Comfortable using AI‑assisted development tools and agentic coding workflows.

Required skills

• Python
• SQL
• AWS
• Databricks
• Spark
• Distributed computing
• Large‑scale storage
• Data modeling
• System design
• Debugging distributed systems
• AI‑assisted development tools