Staff Software Engineer, Data Eval & Simulation

Perplexity AI; United States; On-site; USD 120,000 - 190,000; Full time

Job summary

Perplexity AI is seeking an experienced software engineer to build and maintain the data pipelines powering eval verdicts and product decisions. You will own the flow from raw signals to durable datasets that influence decision-making across the company, and you’ll develop a simulator to replay user interactions for LLMs and VLMs.

You will ensure data trust with monitoring and lineage in a fast, high-impact team.

Qualifications

• 3+ years of software engineering experience shipping production systems.
• Proficiency in Python and SQL with production-grade code.
• Experience with AWS and lakehouse ecosystems like Databricks or Spark.
• Strong fundamentals in data modeling, system design, and debugging distributed systems.
• Experience with big data systems including distributed compute and large-scale storage.
• Familiarity with LLM/VLM interfaces, tokenization, and multimodal payloads.
• Experience with evaluation platforms, experimentation systems, or ML infrastructure.
• Prior work supporting customer-facing products at scale.

Responsibilities

• Build the systems and pipelines that enable Search, Product, and other teams to access eval verdicts.
• Own the evals-to-product loop and turn raw signals into durable datasets.
• Create a robust simulator pipeline to replay user interactions for LLMs and VLMs.
• Maintain data trust via monitoring, lineage, and quality checks.
• Work in a small, high-impact team shaping Perplexity's answer quality.

Skills

Python

SQL

Data modeling

Distributed systems

AWS

Databricks

Spark

Big data

Debugging

ML infra

Data engineering

LLM/VLM interfaces

Evaluation platforms

Customer-facing products

Tools

Databricks

Spark

AWS

Job description

Perplexity AI is seeking an experienced software engineer to build and maintain the data pipelines powering eval verdicts and product decisions. You will own the flow from raw signals to durable datasets that influence decision-making across the company, and you’ll develop a simulator to replay user interactions for LLMs and VLMs.

You will ensure data trust with monitoring and lineage in a fast, high-impact team.