Audio Annotation Analyst

Role overview

Review and label audio, video, and related interaction data to help research teams develop and evaluate systems for fraud detection, identity verification, and synthetic media detection. The work combines careful, repeatable annotation with targeted investigations into unusual interactions and model behavior.

Responsibilities - Apply project-specific criteria to label recordings and interaction data, including fraud indicators, authentication outcomes, and signs of human or synthetic content. - Review existing annotations for inconsistencies and help keep large datasets accurate and usable. - Flag unclear cases, record observations, and work with research colleagues to resolve questions. - Investigate issues such as background noise, degraded audio, gaps in automated phone workflows, or unexpected model results. - Summarize patterns and findings, and handle sensitive material according to privacy and security procedures.

Requirements - Strong listening and observation skills, with close attention to detail across audio and video. - Ability to maintain quality during focused, high-volume review work. - Sound judgment when applying detailed guidance to ambiguous cases, with clear documentation and escalation. - Familiarity with spreadsheets or similar tools for organizing and reviewing data. - Clear written communication, independent work habits, and ability to manage priorities remotely.

Nice to have - Experience in annotation, transcription, quality assurance, or qualitative data review. - Familiarity with fraud, authentication, voice biometrics, contact center workflows, synthetic speech, or manipulated media. - Exposure to annotation platforms, speech analysis, machine learning labeling, or basic SQL.

Benefits and work setup The role is remote-first, with periodic team gatherings. Listed benefits include flexible paid time off, equity, health and dental coverage, parental leave, a phone and internet allowance, and a learning stipend.