Derrick Karanja

AI Evaluation Analyst ( Business Analyst )

turing

Jan 2026 - Current (8 months)

Working as a freelance AI Evaluation Analyst supporting frontier AI research and enterprise systems. • Conduct side-by-side evaluations of AI-generated outputs for quality, accuracy, and relevance • Perform fact-checking and verification using reliable sources • Provide detailed, structured justifications based on evaluation guidelines • Apply standardized scoring frameworks for task completion and performance • Use browser-based tools and digital workflows for analysis and review • Document findings clearly to support model training and improvement

AI Image Quality Evaluator

iMerit Scholars

Jan 2026 - May 2026 (5 months)

Currently evaluating AI-generated images using structured comparison frameworks to assess quality, accuracy, and prompt alignment. Analyze prompts to determine explicit and implicit intent for accurate evaluation Perform side-by-side (SxS) comparisons to select the best image output Apply standardized evaluation guidelines to ensure consistency and high-quality outputs Provide structured preference labels that support training and improvement of generative AI models

Data Annotator

Uber AI Solutions

Nov 2025 - May 2026 (7 months)

- Annotate high-quality 2D images of industrial infrastructure for AI and machine learning systems in the oil and energy sector. - Label and segment assets, using polygon and bounding box techniques. - Utilize the Encord annotation platform for efficient dataset creation, review, and quality control. - Apply detailed semantic segmentation for complex, overlapping, and irregular objects.

3D Data Annotator – Computer Vision

CloudFactory

Oct 2025 - Current (11 months)

• Annotate 3D building and geospatial imagery for computer vision datasets. • Label rooftops, structures, boundaries, and architectural features. • Ensure spatial accuracy and consistency across large datasets.

LLM Agentic Trainer – JSON

turing

Oct 2025 - Nov 2025 (2 months)

Validating and structuring training data for large language models using Python and JSON, supporting agentic AI systems with consistent, high-quality datasets. Develop Python and JSON-based pipelines ensuring data accuracy and workflow consistency. Collaborate with AI researchers to structure function-calling datasets aligned with LLM training needs.

AI Data Annotator | Machine Learning Data Labeling | Quality Control & Instruction Editing

Erud AI

Oct 2025 - Nov 2025 (2 months)

Working as an independent contractor with Erud AI on diverse machine learning data annotation projects. • Annotating, reviewing, and validating data for AI and LLM training datasets. • Utilizing LabelStudio for task management and Labeling • Ensuring high-quality and consistent annotations aligned with AI model requirements.

Data Entry and Verification

CloudFactory

Dec 2024 - Jun 2025 (7 months)

Conducted data entry, verification, and labeling for various AI and research projects, ensuring high accuracy and consistency. Managed large datasets in CSV and Excel, applying logical validation and cleaning for effective downstream use. Developed a strong attention to detail while maintaining productivity in high-volume annotation environments.

Data Annotation Specialist

CloudFactory

Jun 2024 - Jan 2026 (1 year 8 months)

- Performed high-quality 2D and 3D image annotation for machine learning models across multiple industries. - Annotated 2D images using polygon and bounding box techniques for object detection and segmentation tasks. - Labeled aerial/drone imagery by identifying objects airborne objects for drone training systems. - Conducted industrial asset annotation for oil and gas projects using polygon tools.

AI Data Annotator

CVATAI

Sep 2023 - Feb 2025 (1 year 6 months)

Annotated and labeled complex datasets for computer vision projects using CVAT, focusing on object detection, segmentation, and classification. Improved annotation accuracy by 20% by implementing systematic QA reviews and Python-based validation scripts. Collaborated cross-functionally with ML engineers to refine labeling guidelines and ensure datasets met training standards.

Software Developer

Upwork

Aug 2023 - Jan 2026 (2 years 6 months)

Designed and deployed web applications using JavaScript, Django, React, Tailwind CSS, Next.JS and Flutter for clients in the US and internationally. Developed data scraping and automation scripts in Python to streamline client workflows. Delivered projects on time while maintaining clear communication and high client satisfaction ratings.

Data Entry Specialist

Upwork

Apr 2023 - Jan 2026 (2 years 10 months)

Performed high-volume data entry and cleaning, ensuring client datasets were organized and error-free. Used Excel and Python for data analysis and reporting. Built automated scraping tools to collect structured data efficiently.

3D Data Labeling and Autonomous Vehicle Annotator

Remotasks

Nov 2021 - Feb 2024 (2 years 4 months)

Worked on LiDAR, radar, and multi-sensor fusion datasets for self-driving car training and environment modeling. Executed 3D cuboid labeling for vehicles, pedestrians, cyclists, infrastructure, and road elements across thousands of frames. Performed LiDAR ground segmentation, object tracking across sequences, and annotation of static vs. dynamic objects. Utilized advanced annotation platforms such as Remotasks Pro, 3D Tasker, and AV-focused labeling tools. Ensured high data quality through QA validation, guideline adherence, and detailed error correction. Identified edge cases and contributed feedback that improved annotation instructions for complex 3D scenes. Demonstrated strong attention to detail and ability to work under tight de

About

Detail

Timeline

Résumé