Profile
Háron de Souza profile illustration

Háron de Souza

Data Labeling Analyst | Quality Assurance & Team Lead


LOCATION: Dublin, Ireland

STATUS: Eligible for full-time work in Ireland



Education

BSc in Social Communication (Advertising & Marketing)
Salvador University (UNIFACS), Brazil | 2014 - 2018

Certificates

Graphic & Web Design | Prepara Course, Brazil | 2018
Film Directing & Video Editing | Cinearts Film Institution, Brazil | 2017 - 2018


Languages
Brazilian Portuguese: Native English: Fluent (Professional Working Proficiency)
Professional Summary

AI Operations, Data Labeling, Quality Assurance, and Trust & Safety specialist with over 5 years of experience supporting artificial intelligence, machine learning, content moderation, policy enforcement, model optimization, data annotation, and operational quality within global technology environments.

Experienced in Large Language Model (LLM) training and evaluation, high-complexity case review, Root Cause Analysis (RCA), quality monitoring, error trend analysis, performance management, policy compliance, SOP development, stakeholder reporting, escalation management, workflow optimization, and operational issue resolution.

Proven ability to lead quality-focused teams, analyze individual and team performance, identify systemic quality issues, deliver calibration and policy clarification, manage engineering escalations, and collaborate cross-functionally with Product, Engineering, R&D, Operations, and Trust & Safety teams.

Strong analytical judgment, technical problem-solving skills, and experience working in high-volume, SLA-driven, quality-focused, and ambiguity-heavy environments.

Core Skills
AI & Machine Learning Operations
Data LabelingData AnnotationLarge Language Model (LLM) TrainingLLM EvaluationModel OptimizationModel Performance AnalysisGolden Dataset CreationHigh-Complexity Case ReviewMulti-Modal AnnotationMisclassification AnalysisDataset ReviewAnnotation QualityEdge Case IdentificationAI Training Data Quality
Quality Assurance & Performance
Quality AssuranceQuality MonitoringRoot Cause Analysis (RCA)Quality ImprovementError Trend AnalysisFalse Positive & False Negative AnalysisAgent Performance TrackingQuality ProfilingQuality CalibrationCoaching & FeedbackAccuracy MonitoringProcess AuditingDefect TrackingKPI TrackingOperational Reporting
Team Leadership & Operations
Team LeadershipAgent DevelopmentPerformance ManagementWeekly Quality ReviewsTeam CalibrationStakeholder ReportingCross-Functional CollaborationWorkflow ManagementQueue ManagementOperational ExcellenceProcess ImprovementSLA ManagementHigh-Volume Operations
Trust & Safety
Trust & Safety OperationsPolicy EnforcementContent ModerationPolicy ComplianceRisk MitigationSensitive Content ReviewEscalation ManagementPlatform IntegrityDigital Rights OperationsAdvertising Policy ReviewRegulatory CompliancePolicy Interpretation
Process & Technical Operations
SOP DevelopmentProcess StandardizationWorkflow OptimizationTooling Issue ManagementBug ReportingEngineering EscalationIssue TrackingTechnical TroubleshootingStakeholder CoordinationVendor CoordinationContinuous Improvement
Tools & Technologies
Data & Reporting
Google SheetsMicrosoft ExcelLooker StudioSQL (Basic to Intermediate)Microsoft Project
Project & Quality Management
JiraTrelloTestRailCustom Moderation Platforms
Technical & Automation
JavaScript Workflow ScriptingProcess Automation with AIHTMLWordPressTechnical Troubleshooting
Creative Tools
Adobe PhotoshopAdobe IllustratorAdobe Premiere ProAdobe After EffectsCapCutFilmoraAI Image GenerationCustom AI Model Training Platforms
Professional Experience
Data Labeling Analyst | Quality Assurance & Team Lead | Meta (via Morgan McKinley)

Dublin, Ireland | Jan 2026 - Present

  • Promoted into a Quality Assurance and team leadership role with responsibility for the weekly quality performance, development, and day-to-day support of a dedicated group of 7 Data Labeling Analysts, while acting as the primary point of contact for policy, workflow, and tooling queries, supporting agents with unclear cases, operational concerns, and issues affecting their ability to complete reviews accurately.
  • Perform reviews on a specialized, high-complexity queue, acting as an authoritative decision-maker on cases requiring definitive resolution and contributing validated decisions to golden datasets used for LLM training and model quality improvement.
  • Conduct weekly Root Cause Analysis (RCA) reviews of work completed by both Data Labeling Analysts and Senior Expert Data Labeling Analysts, identifying confirmed errors, recurring mistake trends, policy knowledge gaps, false positive and false negative tendencies, and performance patterns across policy categories. Maintain individual quality profiles, track improvements and recurring weaknesses, and use findings to identify coaching, calibration, training, and broader process-improvement opportunities.
  • Deliver individual quality feedback and lead weekly team meetings focused on policy updates, trending mistakes, complex cases, and shared learnings from errors identified across both the assigned group and the wider operation.
  • Produce and present weekly quality reports to management, summarizing RCA findings, individual and team performance, policy-category trends, development needs, emerging risks, quality profiles, and measurable improvements over time.
  • Own the reporting and follow-up of tooling issues and operational bugs across the wider team, creating engineering tickets, documenting reproduction steps and supporting evidence, coordinating directly with technical stakeholders, and tracking issues through investigation, resolution, and closure.
Data Labeling Analyst | Meta (via Morgan McKinley)

Dublin, Ireland | Jan 2025 - Jan 2026

  • Labeled, reviewed, and analyzed large-scale datasets used to support artificial intelligence, machine learning, and platform integrity initiatives.
  • Trained, evaluated, and calibrated Large Language Models (LLMs) by reviewing model outputs, correcting misclassifications, and applying detailed annotation guidelines.
  • Identified edge cases, labeling inconsistencies, policy gaps, and systemic quality issues affecting model performance and review accuracy.
  • Performed quality analysis and investigated recurring error patterns to support improvements in annotation consistency and model behavior.
  • Escalated complex cases and structural issues requiring additional review, policy clarification, or operational guidance.
  • Collaborated cross-functionally with Product, Engineering, Operations, and Trust & Safety teams to communicate operational findings and support workflow and model improvements.
  • Maintained high levels of accuracy and productivity while working within strict quality, volume, and turnaround requirements.
  • Designed and implemented a lightweight JavaScript bookmarklet to improve URL and domain verification during review workflows, reducing repetitive manual verification steps and supporting improved review accuracy and average handling time.
Data Understanding Specialist | TikTok

Dublin, Ireland | Jan 2024 - Jan 2025

  • Managed high-priority seasonal, Tier 3, and complex operational queues in fast-paced and ambiguity-driven environments.
  • Conducted Root Cause Analysis (RCA) to identify causes of model performance degradation, review inconsistencies, and operational quality issues.
  • Analyzed labeling errors and operational trends to identify recurring problems and improvement opportunities.
  • Partnered with Product Management, Research and Development, and operational stakeholders to improve labeling frameworks and AI model performance.
  • Developed, reviewed, and maintained Standard Operating Procedures (SOPs) for complex, seasonal, and high-priority workflows.
  • Improved process consistency by clarifying decision criteria, operational procedures, and labeling expectations.
  • Led recurring escalation and policy alignment discussions with cross-functional stakeholders.
  • Supported policy interpretation and resolution of ambiguous or complex review scenarios.
  • Consolidated 12 concurrent operational queues into 8 optimized workflows, reducing process overlap and improving seasonal project governance, queue management, and performance tracking.
Model Optimization Specialist | TikTok

Dublin, Ireland | Apr 2022 - Jan 2024

  • Reviewed advertising content for compliance with platform policies, advertising standards, and integrity requirements.
  • Labeled and classified multimedia content according to machine learning training specifications.
  • Supported AI and machine learning model optimization through accurate annotation and classification.
  • Analyzed operational dashboards, quality reports, and performance data to identify trends and optimization opportunities.
  • Investigated model and reviewer inconsistencies affecting content classification and policy enforcement.
  • Coordinated cross-functional investigations involving escalated operational and monetization integrity issues.
  • Supported accurate policy enforcement in high-volume and performance-driven operational environments.
Digital Rights Operations Agent | Meta (via Accenture)

Dublin, Ireland | Jul 2021 - Apr 2022

  • Investigated flagged online content to determine compliance with platform policies, intellectual property requirements, and regulatory standards.
  • Reviewed sensitive and high-risk content requiring detailed policy interpretation and operational judgment.
  • Managed urgent escalations and complex enforcement cases.
  • Identified suspicious activity, behavioral trends, and recurring patterns to support platform integrity decisions.
  • Collaborated with global stakeholders on complex content, integrity, and enforcement cases.
  • Maintained accuracy and productivity requirements within SLA-driven operational workflows.
Game Quality Assurance - LQA & FQA | Keywords Studios

Dublin, Ireland | Jun 2019 - Jul 2021

  • Performed Linguistic Quality Assurance (LQA) and Functional Quality Assurance (FQA) testing on AAA video game titles.
  • Tested game content for localization quality, linguistic accuracy, functionality, usability, and platform compliance.
  • Identified, documented, reproduced, and tracked software and localization defects.
  • Logged detailed bug reports using Jira and TestRail.
  • Collaborated with development and QA teams to investigate and resolve defects.
  • Performed regression testing and validation of fixes.
  • Supported delivery of high-quality localized gaming experiences across multiple projects.

Selected AAA Game Credits
Assassin's Creed Discovery Tour: Ancient GreeceDreams - Sony Interactive EntertainmentA Total War Saga: TroyResident Evil VillageRatchet & Clank: Rift ApartLemnis GateGhost of Tsushima: Director's Cut
Earlier Professional Experience

Social Media Manager (Self-Employed)

Brazil | 2018

  • Managed social media content, digital communications, and audience engagement.
  • Created visual and written content for digital platforms.

Graphic Designer & Social Media Manager | UP Media

Brazil | 2018

  • Produced graphic design assets and social media content.
  • Supported digital marketing and online brand communication activities.

Sales Development Representative | WebFoco Telecom

Brazil | 2016 - 2018

  • Conducted outbound prospecting and customer acquisition activities.
  • Communicated with prospective customers and supported sales pipeline development.

Outbound Call Center Agent | Contax Contact Center

Brazil | 2014 - 2016

  • Conducted outbound customer communication within target-driven contact center environments.
  • Managed customer interactions while meeting productivity and service quality requirements.