Principal Machine Learning Engineer

Overview As a Principal Machine Learning Engineer at ZoomInfo, you set the technical direction for high-impact ML problems that enhance our data graph and AI agents. You own end-to-end outcomes, choosing the right mix of classical ML, statistics, and language models to scale and improve data quality, entity resolution, and buying-intent inference. You lead design reviews and mentor others, shaping solutions that run at web-scale while balancing cost and accuracy. This role offers hands-on impact in a fast, collaborative environment. Compensation / Benefitscomprehensive benefitsholistic mind, body and lifestyle programsbase salary disclosurebonus, commission, equity and other benefits may applyremote-friendly work (#LI-Remote)career development and mentorship opportunities ResponsibilitiesExtend the data graph into long-tail companies by extracting leadership, locations, and products from websites and detecting stale recordsPredict unknown attributes (headcount, revenue) for under-documented companies using gradient-boosted trees and regression with missing inputsDetermine if two records describe the same company or person, optimizing for correct merges and splitsInfer buying intent from multilingual, noisy text to support propensity scoring, lookalike retrieval, and contact recommendationsBuild agents that research companies with sourced citations and design robust evaluations to separate correct results from superficial completionsDistill large models into cost-effective, smaller versions and manage quantization within the same workflowLead ambiguous, high-stakes problems from concept to shipped solutions, and raise engineering standards through reviews and mentorship Key requirementsProduction ML systems ownership with technical leadership as an individual contributorClassical ML expertise plus feature engineering on large, messy tabular data and applied statisticsExperience with language processing at scale (text classification, information extraction, entity linking)Experience building and operating LLM agents or multi-step systems in productionProficiency in Python and SQL with distributed data processing; comfort with AI coding tools and rigorous output reviewleadershipmentorshipdesign reviewsclassical ML and statisticsfeature engineering on large tabular datatext processing at scale (classification, information extraction, entity linking)