Senior Software Engineer, Pretraining Data

Quick Facts

• Role: Senior Software Engineer, Pretraining Data

Description

Build and own the data systems behind frontier model initial training. You will design large-scale crawling, data platform, and pipeline infrastructure that converts raw web, code, multimodal, and acquired data into training-ready datasets. You’ll set technical direction for these systems, lead cross-team projects, and mentor engineers as pretraining data scales.

Responsibilities

• Data Platform: Design, build, and scale distributed infrastructure and high-throughput data pipelines to transform raw data into pretraining datasets with orchestration, tooling, and end-to-end visibility into lineage, freshness, and pipeline health.

• Data Quality: Build and deploy high-performance models/algorithms to classify, rank, deduplicate, filter, and clean training data; run experiments on data mixtures, sampling strategies, and quality thresholds to measure impact on training loss and downstream evaluations.

• Crawling: Build and optimize large-scale web crawling for URL discovery, prioritization, host scheduling, fetching, parsing, and extraction; improve coverage, efficiency, and reliability with intelligent scheduling, anti-bot resilience, robust extraction, and automated recovery.

• Technical Leadership & Collaboration: Lead complex initiatives across data and model-training teams, influence architecture, mentor engineers, and translate improvements into measurable gains in model performance.

Requirements

• 5+ years of professional software engineering experience building distributed systems and/or large-scale data infrastructure.

• Onsite role: must be able to work in person at the Palo Alto office.

Benefits

• Annual salary range: $220,000–$357,000 USD (level and base salary determined case-by-case).

• Equity.

• Comprehensive medical, vision, and dental coverage.

• Access to a 401(k) retirement plan.

• Short- and long-term disability insurance and life insurance.

• Various other discounts and perks.