JOBSEARCHER

Applied Scientist, Data Processing Agents Science

AmazonSan Jose, CAL6 LeadSeptember 15th, 2026
Overview In this role you apply agentic AI/ML and systems expertise to improve large-scale analytics services. You will design and build autonomous AI-driven solutions to optimize big data workflows and collaborate across engineering, product, and business teams to put innovations into production at AWS scale. You’ll benchmark science implementations, publish findings, and mentor peers, all while advancing transformative, science-led experiences for customers. This position offers a chance to shape scalable, data-driven AI solutions within a collaborative, inclusive culture at AWS. Compensation / Benefitshealth insurance (medical, dental, vision, prescription)401(k) matchingRSUspaid time offparental leaveflexible work hours ResponsibilitiesApply agentic AI/ML and systems skills to impact large analytics servicesDesign, build, and benchmark agentic solutions and ML-powered automation for big data workflowsCollaborate with applied science peers and with engineering, product, and business leaders to launch work in production at scaleDefine science features with product teams and customers, create science plans, and implement state-of-the-art solutionsPrepare, publish, and present scientific work at top-tier venues and evangelize findings internallyMentor applied science interns and teammates, participate in knowledge-sharing activitiesStay current with research through reading groups and external/internal seminars Key requirementsPhD, or Master’s degree with 4+ years in CS, CE, ML or related fieldExperience in patents or publications at top-tier conferences or journalsProgramming experience in Java, C++, Python or related languages3+ years researching agentic AI or ML techniques applied to systems problems in programming languages, databases, or distributed systemscollaborationmentorshipeffective communicationagentic AI/MLsystems-focused ML for databases, programming languages, and distributed systemsbenchmarking and evaluating ML implementations