Jane Maina

Key Skills

Software

Internal/Proprietary Tooling

Labelbox

Top Subject Matter

No subject matter listed

Top Data Types

Image

Text

Geospatial Tiled Imagery

Top Label Types

Classification

Text Generation

Emotion Recognition

Object Detection

Mapping

Question Answering

Action Recognition

Land Cover Classification

Evaluation Rating

Computer Programming Coding

Data Collection

Freelancer Overview

I am a highly experienced GIS and remote sensing specialist with a strong background in data annotation, labeling, and AI training data for computer vision and geospatial applications. My work includes labeling and categorizing imagery, video, audio, and text data from satellite, aerial, drone, and LiDAR sources, ensuring precise tagging and quality for AI model development. I have hands-on experience with tools such as ArcGIS, QGIS, ENVI, Google Earth Engine, and PostgreSQL/PostGIS, and I am proficient in Python and SQL for automating data pipelines and quality assurance. I have contributed to large-scale projects, including national GIS data collection, emergency infrastructure planning, and AI workflow evaluation, where I validated outputs, documented edge cases, and ensured strict compliance with annotation guidelines. My attention to detail, technical expertise, and collaborative approach ensure high-quality, reliable datasets that drive robust AI and machine learning solutions.

Labeling Experience

Data Annotator – Computer Use Agent (CUA) Evaluation

Internal Proprietary ToolingGeospatial Tiled ImageryMappingLand Cover Classification

• Execute structured end-to-end evaluations of AI Computer Use Agent workflows across desktop applications (e.g., QGIS, OnlyOffice, Calibre, Notepad++, system utilities), ensuring strict adherence to Standard Operating Procedures (SOPs). • Validate task execution accuracy by identifying hard blockers, tool limitations, system restrictions, and deviations from prompt specifications in controlled test environments. • Document execution paths, edge cases, and failure points with clear, reproducible reporting to improve model reliability, procedural compliance, and real-world usability. • Enforce compliance standards by confirming that outputs meet exact prompt requirements without unauthorized substitutions, workarounds, or assumption-based deviations.

• Execute structured end-to-end evaluations of AI Computer Use Agent workflows across desktop applications (e.g., QGIS, OnlyOffice, Calibre, Notepad++, system utilities), ensuring strict adherence to Standard Operating Procedures (SOPs). • Validate task execution accuracy by identifying hard blockers, tool limitations, system restrictions, and deviations from prompt specifications in controlled test environments. • Document execution paths, edge cases, and failure points with clear, reproducible reporting to improve model reliability, procedural compliance, and real-world usability. • Enforce compliance standards by confirming that outputs meet exact prompt requirements without unauthorized substitutions, workarounds, or assumption-based deviations.

2025

QA – AI Model Training (Persuasion)

LabelboxTextText GenerationAction Recognition

• Perform quality assurance on AI training outputs by reviewing annotations, arguments, and model responses for accuracy, logical consistency, and adherence to project guidelines. • Evaluate contributor work to ensure balanced reasoning, ethical persuasion, and high linguistic quality before inclusion in AI training datasets. • Identify errors and inconsistencies, provide structured feedback, and recommend improvements to QA standards and evaluation guidelines to enhance model performance.

• Perform quality assurance on AI training outputs by reviewing annotations, arguments, and model responses for accuracy, logical consistency, and adherence to project guidelines. • Evaluate contributor work to ensure balanced reasoning, ethical persuasion, and high linguistic quality before inclusion in AI training datasets. • Identify errors and inconsistencies, provide structured feedback, and recommend improvements to QA standards and evaluation guidelines to enhance model performance.

2025 - 2025

Persuasion / AI Model Trainer

LabelboxTextQuestion AnsweringText Generation

• Analyze and structure arguments: Review assigned topics and create balanced lists of reasons for/against a statement to train AI models in persuasive reasoning. • Support model training and evaluation: Assist senior analysts by summarizing content, checking quality, and ensuring the AI understands logical and persuasive patterns. • Communicate insights clearly: Provide concise written or spoken explanations in English to help refine AI decision-making and argument generation.

• Analyze and structure arguments: Review assigned topics and create balanced lists of reasons for/against a statement to train AI models in persuasive reasoning. • Support model training and evaluation: Assist senior analysts by summarizing content, checking quality, and ensuring the AI understands logical and persuasive patterns. • Communicate insights clearly: Provide concise written or spoken explanations in English to help refine AI decision-making and argument generation.

2025 - 2025

Data Annotator

Internal Proprietary ToolingImageClassificationText Generation

• Labelled and categorized imagery, video, audio, and text data for various annotation projects. • Worked with datasets from satellite, aerial, drone, and LiDAR sources, ensuring precise tagging and visual identification of features. • Followed detailed annotation guidelines to maintain accuracy, consistency, and quality across all labeled datasets. • Demonstrated strong attention to detail and reliability while managing high volumes of visual and audio data. • Contributed to producing high-quality annotated data supporting AI model training.

• Labelled and categorized imagery, video, audio, and text data for various annotation projects. • Worked with datasets from satellite, aerial, drone, and LiDAR sources, ensuring precise tagging and visual identification of features. • Followed detailed annotation guidelines to maintain accuracy, consistency, and quality across all labeled datasets. • Demonstrated strong attention to detail and reliability while managing high volumes of visual and audio data. • Contributed to producing high-quality annotated data supporting AI model training.

2025 - 2025

Education

Work History