Dancan Wanyonyi

Key Skills

Software

Scale AI

Remotasks

CrowdSource

Telus

Appen

CloudFactory

Clickworker

Data Annotation Tech

Top Subject Matter

Conversational AI

Llms Domain Expertise

Prompt Evaluation

Top Data Types

Text

Image

Document

Top Task Types

RLHF

Bounding Box

Object Detection

Segmentation

Text Generation

Text Summarization

Data Collection

Transcription

Prompt Response Writing SFT

Evaluation Rating

Freelancer Overview

LLM Evaluator. Brings 7+ years of professional experience across legal operations, contract review, compliance, and structured analysis. Core strengths include Scale AI and Remotasks. Education includes Bachelor of Arts, University of Nairobi (2020). AI-training focus includes data types such as Text and Image and labeling workflows including RLHF, Bounding Box, and Object Detection.

Labeling Experience

LLM Evaluator

Scale AITextRLHF

As an LLM Evaluator, I contributed to the training and evaluation of large language models. My work involved labeling and categorizing text data, ranking model outputs, and performing quality assurance for RLHF pipelines. I developed prompt datasets to enhance the performance and safety of conversational AI systems. • Conducted LLM prompt evaluation for reasoning, bias, and hallucinations. • Labeled, classified, and ranked large volumes of text outputs. • Performed multi-stage QA checks to maintain accuracy above 98%. • Created and curated datasets to support LLM fine-tuning and evaluation.

As an LLM Evaluator, I contributed to the training and evaluation of large language models. My work involved labeling and categorizing text data, ranking model outputs, and performing quality assurance for RLHF pipelines. I developed prompt datasets to enhance the performance and safety of conversational AI systems. • Conducted LLM prompt evaluation for reasoning, bias, and hallucinations. • Labeled, classified, and ranked large volumes of text outputs. • Performed multi-stage QA checks to maintain accuracy above 98%. • Created and curated datasets to support LLM fine-tuning and evaluation.

2020 - Present

Data Annotation Specialist – Autonomous Systems

RemotasksImageBounding Box

As a Data Annotation Specialist in autonomous systems, I annotated diverse driving datasets to train perception models. My tasks included bounding box annotation, segmentation, tracking across frames, and LiDAR point cloud labeling. I focused on improving dataset quality and provided feedback to optimize annotation workflows. • Annotated vehicles, pedestrians, and traffic signs using bounding boxes. • Executed semantic segmentation and object tracking for road scenes. • Labeled 3D LiDAR point clouds for spatial understanding. • Helped reduce mislabeling errors and increased annotation throughput.

As a Data Annotation Specialist in autonomous systems, I annotated diverse driving datasets to train perception models. My tasks included bounding box annotation, segmentation, tracking across frames, and LiDAR point cloud labeling. I focused on improving dataset quality and provided feedback to optimize annotation workflows. • Annotated vehicles, pedestrians, and traffic signs using bounding boxes. • Executed semantic segmentation and object tracking for road scenes. • Labeled 3D LiDAR point clouds for spatial understanding. • Helped reduce mislabeling errors and increased annotation throughput.

2017 - 2019

LLM Prompt Evaluation & Safety Testing

Scale AIText

For the LLM Prompt Evaluation & Safety Testing project, I performed safety scoring and qualitative evaluations on model-generated outputs. The responsibilities included judging for toxicity, bias, factuality, and reasoning quality to guide LLM improvements. I also developed prompts to test models in real-world scenarios and assess conversational coherence. • Scored model outputs for various safety and reasoning metrics. • Developed and utilized 300+ prompts for evaluation. • Contributed to safer, more accurate LLM deployment. • Assessed conversational capabilities in diverse contexts.

For the LLM Prompt Evaluation & Safety Testing project, I performed safety scoring and qualitative evaluations on model-generated outputs. The responsibilities included judging for toxicity, bias, factuality, and reasoning quality to guide LLM improvements. I also developed prompts to test models in real-world scenarios and assess conversational coherence. • Scored model outputs for various safety and reasoning metrics. • Developed and utilized 300+ prompts for evaluation. • Contributed to safer, more accurate LLM deployment. • Assessed conversational capabilities in diverse contexts.

Not specified

Autonomous Driving Dataset Labeling Project

RemotasksImageObject Detection

During an autonomous driving dataset labeling project, I contributed to annotating over 10,000 frames for perception models. The project involved identifying and labeling objects, lanes, and road features to support model training. My work helped improve object detection accuracy for autonomous systems. • Labeled large-scale autonomous driving data for machine learning. • Focused on improving object, lane, and road feature detection. • Supported accurate perception model development for safety applications. • Contributed to a dataset used for object detection performance gains.

During an autonomous driving dataset labeling project, I contributed to annotating over 10,000 frames for perception models. The project involved identifying and labeling objects, lanes, and road features to support model training. My work helped improve object detection accuracy for autonomous systems. • Labeled large-scale autonomous driving data for machine learning. • Focused on improving object, lane, and road feature detection. • Supported accurate perception model development for safety applications. • Contributed to a dataset used for object detection performance gains.

Not specified

Education

Work History